{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.879128727013033,
  "eval_steps": 100,
  "global_step": 41500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00023805272867940248,
      "grad_norm": 3.546875,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.1991,
      "step": 1
    },
    {
      "epoch": 0.00047610545735880496,
      "grad_norm": 3.046875,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.3933,
      "step": 2
    },
    {
      "epoch": 0.0007141581860382074,
      "grad_norm": 3.96875,
      "learning_rate": 1.8e-06,
      "loss": 2.0374,
      "step": 3
    },
    {
      "epoch": 0.0009522109147176099,
      "grad_norm": 3.515625,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 2.275,
      "step": 4
    },
    {
      "epoch": 0.0011902636433970125,
      "grad_norm": 3.375,
      "learning_rate": 3e-06,
      "loss": 2.7338,
      "step": 5
    },
    {
      "epoch": 0.0014283163720764148,
      "grad_norm": 3.5,
      "learning_rate": 3.6e-06,
      "loss": 2.2785,
      "step": 6
    },
    {
      "epoch": 0.0016663691007558173,
      "grad_norm": 3.171875,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 2.5345,
      "step": 7
    },
    {
      "epoch": 0.0019044218294352198,
      "grad_norm": 3.015625,
      "learning_rate": 4.800000000000001e-06,
      "loss": 2.1867,
      "step": 8
    },
    {
      "epoch": 0.0021424745581146223,
      "grad_norm": 3.28125,
      "learning_rate": 5.4e-06,
      "loss": 2.0473,
      "step": 9
    },
    {
      "epoch": 0.002380527286794025,
      "grad_norm": 2.984375,
      "learning_rate": 6e-06,
      "loss": 2.2222,
      "step": 10
    },
    {
      "epoch": 0.0026185800154734274,
      "grad_norm": 3.453125,
      "learning_rate": 6.6e-06,
      "loss": 2.3704,
      "step": 11
    },
    {
      "epoch": 0.0028566327441528296,
      "grad_norm": 3.3125,
      "learning_rate": 7.2e-06,
      "loss": 2.475,
      "step": 12
    },
    {
      "epoch": 0.0030946854728322324,
      "grad_norm": 3.171875,
      "learning_rate": 7.8e-06,
      "loss": 2.3317,
      "step": 13
    },
    {
      "epoch": 0.0033327382015116347,
      "grad_norm": 3.28125,
      "learning_rate": 8.400000000000001e-06,
      "loss": 1.9672,
      "step": 14
    },
    {
      "epoch": 0.0035707909301910374,
      "grad_norm": 3.171875,
      "learning_rate": 9e-06,
      "loss": 2.3329,
      "step": 15
    },
    {
      "epoch": 0.0038088436588704397,
      "grad_norm": 3.453125,
      "learning_rate": 9.600000000000001e-06,
      "loss": 2.2018,
      "step": 16
    },
    {
      "epoch": 0.004046896387549842,
      "grad_norm": 3.5,
      "learning_rate": 1.02e-05,
      "loss": 2.2174,
      "step": 17
    },
    {
      "epoch": 0.004284949116229245,
      "grad_norm": 4.03125,
      "learning_rate": 1.08e-05,
      "loss": 2.1781,
      "step": 18
    },
    {
      "epoch": 0.004523001844908647,
      "grad_norm": 2.890625,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 2.1976,
      "step": 19
    },
    {
      "epoch": 0.00476105457358805,
      "grad_norm": 2.90625,
      "learning_rate": 1.2e-05,
      "loss": 2.1289,
      "step": 20
    },
    {
      "epoch": 0.004999107302267452,
      "grad_norm": 3.671875,
      "learning_rate": 1.26e-05,
      "loss": 2.2965,
      "step": 21
    },
    {
      "epoch": 0.005237160030946855,
      "grad_norm": 2.96875,
      "learning_rate": 1.32e-05,
      "loss": 2.1637,
      "step": 22
    },
    {
      "epoch": 0.005475212759626257,
      "grad_norm": 2.734375,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 2.0862,
      "step": 23
    },
    {
      "epoch": 0.005713265488305659,
      "grad_norm": 3.15625,
      "learning_rate": 1.44e-05,
      "loss": 2.3559,
      "step": 24
    },
    {
      "epoch": 0.0059513182169850624,
      "grad_norm": 2.984375,
      "learning_rate": 1.5e-05,
      "loss": 2.2741,
      "step": 25
    },
    {
      "epoch": 0.006189370945664465,
      "grad_norm": 3.234375,
      "learning_rate": 1.56e-05,
      "loss": 2.062,
      "step": 26
    },
    {
      "epoch": 0.006427423674343867,
      "grad_norm": 3.53125,
      "learning_rate": 1.62e-05,
      "loss": 2.439,
      "step": 27
    },
    {
      "epoch": 0.006665476403023269,
      "grad_norm": 3.0625,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.8846,
      "step": 28
    },
    {
      "epoch": 0.0069035291317026725,
      "grad_norm": 2.84375,
      "learning_rate": 1.74e-05,
      "loss": 2.1722,
      "step": 29
    },
    {
      "epoch": 0.007141581860382075,
      "grad_norm": 3.515625,
      "learning_rate": 1.8e-05,
      "loss": 2.3224,
      "step": 30
    },
    {
      "epoch": 0.007379634589061477,
      "grad_norm": 2.8125,
      "learning_rate": 1.86e-05,
      "loss": 2.304,
      "step": 31
    },
    {
      "epoch": 0.007617687317740879,
      "grad_norm": 3.265625,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.1407,
      "step": 32
    },
    {
      "epoch": 0.007855740046420282,
      "grad_norm": 2.921875,
      "learning_rate": 1.98e-05,
      "loss": 2.057,
      "step": 33
    },
    {
      "epoch": 0.008093792775099685,
      "grad_norm": 3.765625,
      "learning_rate": 2.04e-05,
      "loss": 2.2785,
      "step": 34
    },
    {
      "epoch": 0.008331845503779087,
      "grad_norm": 3.484375,
      "learning_rate": 2.1e-05,
      "loss": 2.3002,
      "step": 35
    },
    {
      "epoch": 0.00856989823245849,
      "grad_norm": 3.140625,
      "learning_rate": 2.16e-05,
      "loss": 2.224,
      "step": 36
    },
    {
      "epoch": 0.008807950961137892,
      "grad_norm": 3.59375,
      "learning_rate": 2.22e-05,
      "loss": 2.4052,
      "step": 37
    },
    {
      "epoch": 0.009046003689817294,
      "grad_norm": 3.484375,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 2.4158,
      "step": 38
    },
    {
      "epoch": 0.009284056418496696,
      "grad_norm": 2.609375,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 1.9191,
      "step": 39
    },
    {
      "epoch": 0.0095221091471761,
      "grad_norm": 3.5625,
      "learning_rate": 2.4e-05,
      "loss": 2.4574,
      "step": 40
    },
    {
      "epoch": 0.009760161875855503,
      "grad_norm": 3.21875,
      "learning_rate": 2.4599999999999998e-05,
      "loss": 2.1637,
      "step": 41
    },
    {
      "epoch": 0.009998214604534905,
      "grad_norm": 4.15625,
      "learning_rate": 2.52e-05,
      "loss": 2.2556,
      "step": 42
    },
    {
      "epoch": 0.010236267333214307,
      "grad_norm": 3.046875,
      "learning_rate": 2.58e-05,
      "loss": 2.069,
      "step": 43
    },
    {
      "epoch": 0.01047432006189371,
      "grad_norm": 3.28125,
      "learning_rate": 2.64e-05,
      "loss": 2.3573,
      "step": 44
    },
    {
      "epoch": 0.010712372790573112,
      "grad_norm": 3.375,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 2.4745,
      "step": 45
    },
    {
      "epoch": 0.010950425519252514,
      "grad_norm": 3.421875,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 2.3161,
      "step": 46
    },
    {
      "epoch": 0.011188478247931916,
      "grad_norm": 4.21875,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 2.1699,
      "step": 47
    },
    {
      "epoch": 0.011426530976611319,
      "grad_norm": 3.40625,
      "learning_rate": 2.88e-05,
      "loss": 2.0649,
      "step": 48
    },
    {
      "epoch": 0.011664583705290723,
      "grad_norm": 3.015625,
      "learning_rate": 2.94e-05,
      "loss": 2.277,
      "step": 49
    },
    {
      "epoch": 0.011902636433970125,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.6407,
      "step": 50
    },
    {
      "epoch": 0.012140689162649527,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.5895,
      "step": 51
    },
    {
      "epoch": 0.01237874189132893,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 52
    },
    {
      "epoch": 0.012616794620008332,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 53
    },
    {
      "epoch": 0.012854847348687734,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 54
    },
    {
      "epoch": 0.013092900077367136,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.5332,
      "step": 55
    },
    {
      "epoch": 0.013330952806046539,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 56
    },
    {
      "epoch": 0.013569005534725943,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 57
    },
    {
      "epoch": 0.013807058263405345,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 58
    },
    {
      "epoch": 0.014045110992084747,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 59
    },
    {
      "epoch": 0.01428316372076415,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3023,
      "step": 60
    },
    {
      "epoch": 0.014521216449443552,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 61
    },
    {
      "epoch": 0.014759269178122954,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 62
    },
    {
      "epoch": 0.014997321906802356,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 63
    },
    {
      "epoch": 0.015235374635481759,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 64
    },
    {
      "epoch": 0.015473427364161161,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 65
    },
    {
      "epoch": 0.015711480092840565,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 66
    },
    {
      "epoch": 0.015949532821519966,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 67
    },
    {
      "epoch": 0.01618758555019937,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 68
    },
    {
      "epoch": 0.01642563827887877,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4308,
      "step": 69
    },
    {
      "epoch": 0.016663691007558174,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 70
    },
    {
      "epoch": 0.016901743736237578,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3828,
      "step": 71
    },
    {
      "epoch": 0.01713979646491698,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 72
    },
    {
      "epoch": 0.017377849193596383,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.5115,
      "step": 73
    },
    {
      "epoch": 0.017615901922275783,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 74
    },
    {
      "epoch": 0.017853954650955187,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.5325,
      "step": 75
    },
    {
      "epoch": 0.018092007379634588,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 76
    },
    {
      "epoch": 0.018330060108313992,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 77
    },
    {
      "epoch": 0.018568112836993392,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 78
    },
    {
      "epoch": 0.018806165565672796,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 79
    },
    {
      "epoch": 0.0190442182943522,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 80
    },
    {
      "epoch": 0.0192822710230316,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 81
    },
    {
      "epoch": 0.019520323751711005,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 82
    },
    {
      "epoch": 0.019758376480390406,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4151,
      "step": 83
    },
    {
      "epoch": 0.01999642920906981,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 84
    },
    {
      "epoch": 0.02023448193774921,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 85
    },
    {
      "epoch": 0.020472534666428614,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 86
    },
    {
      "epoch": 0.020710587395108015,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 87
    },
    {
      "epoch": 0.02094864012378742,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 88
    },
    {
      "epoch": 0.021186692852466823,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 89
    },
    {
      "epoch": 0.021424745581146223,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 90
    },
    {
      "epoch": 0.021662798309825627,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 91
    },
    {
      "epoch": 0.021900851038505028,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 92
    },
    {
      "epoch": 0.022138903767184432,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3022,
      "step": 93
    },
    {
      "epoch": 0.022376956495863833,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 94
    },
    {
      "epoch": 0.022615009224543237,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 95
    },
    {
      "epoch": 0.022853061953222637,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 96
    },
    {
      "epoch": 0.02309111468190204,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 97
    },
    {
      "epoch": 0.023329167410581445,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 98
    },
    {
      "epoch": 0.023567220139260846,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 99
    },
    {
      "epoch": 0.02380527286794025,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 100
    },
    {
      "epoch": 0.02380527286794025,
      "eval_runtime": 0.5348,
      "eval_samples_per_second": 1.87,
      "eval_steps_per_second": 1.87,
      "step": 100
    },
    {
      "epoch": 0.02404332559661965,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4156,
      "step": 101
    },
    {
      "epoch": 0.024281378325299054,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 102
    },
    {
      "epoch": 0.024519431053978455,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 103
    },
    {
      "epoch": 0.02475748378265786,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 104
    },
    {
      "epoch": 0.02499553651133726,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 105
    },
    {
      "epoch": 0.025233589240016664,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 106
    },
    {
      "epoch": 0.025471641968696068,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 107
    },
    {
      "epoch": 0.025709694697375468,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 108
    },
    {
      "epoch": 0.025947747426054872,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 109
    },
    {
      "epoch": 0.026185800154734273,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4303,
      "step": 110
    },
    {
      "epoch": 0.026423852883413677,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.7162,
      "step": 111
    },
    {
      "epoch": 0.026661905612093077,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 112
    },
    {
      "epoch": 0.02689995834077248,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 113
    },
    {
      "epoch": 0.027138011069451885,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 114
    },
    {
      "epoch": 0.027376063798131286,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3631,
      "step": 115
    },
    {
      "epoch": 0.02761411652681069,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 116
    },
    {
      "epoch": 0.02785216925549009,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 117
    },
    {
      "epoch": 0.028090221984169494,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 118
    },
    {
      "epoch": 0.028328274712848895,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 119
    },
    {
      "epoch": 0.0285663274415283,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 120
    },
    {
      "epoch": 0.0288043801702077,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 121
    },
    {
      "epoch": 0.029042432898887104,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 122
    },
    {
      "epoch": 0.029280485627566508,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 123
    },
    {
      "epoch": 0.029518538356245908,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2454,
      "step": 124
    },
    {
      "epoch": 0.029756591084925312,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3492,
      "step": 125
    },
    {
      "epoch": 0.029994643813604713,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 126
    },
    {
      "epoch": 0.030232696542284117,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 127
    },
    {
      "epoch": 0.030470749270963517,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 128
    },
    {
      "epoch": 0.03070880199964292,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 129
    },
    {
      "epoch": 0.030946854728322322,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 130
    },
    {
      "epoch": 0.031184907457001726,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2603,
      "step": 131
    },
    {
      "epoch": 0.03142296018568113,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 132
    },
    {
      "epoch": 0.031661012914360534,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5031,
      "step": 133
    },
    {
      "epoch": 0.03189906564303993,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3436,
      "step": 134
    },
    {
      "epoch": 0.032137118371719335,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3235,
      "step": 135
    },
    {
      "epoch": 0.03237517110039874,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 136
    },
    {
      "epoch": 0.03261322382907814,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8328,
      "step": 137
    },
    {
      "epoch": 0.03285127655775754,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 138
    },
    {
      "epoch": 0.033089329286436944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 139
    },
    {
      "epoch": 0.03332738201511635,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 140
    },
    {
      "epoch": 0.03356543474379575,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 141
    },
    {
      "epoch": 0.033803487472475156,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 142
    },
    {
      "epoch": 0.03404154020115455,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 143
    },
    {
      "epoch": 0.03427959292983396,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 144
    },
    {
      "epoch": 0.03451764565851336,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 145
    },
    {
      "epoch": 0.034755698387192765,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 146
    },
    {
      "epoch": 0.03499375111587216,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 147
    },
    {
      "epoch": 0.03523180384455157,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 148
    },
    {
      "epoch": 0.03546985657323097,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 149
    },
    {
      "epoch": 0.035707909301910375,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.5769,
      "step": 150
    },
    {
      "epoch": 0.03594596203058978,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 151
    },
    {
      "epoch": 0.036184014759269176,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 152
    },
    {
      "epoch": 0.03642206748794858,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 153
    },
    {
      "epoch": 0.036660120216627984,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 154
    },
    {
      "epoch": 0.03689817294530739,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 155
    },
    {
      "epoch": 0.037136225673986785,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 156
    },
    {
      "epoch": 0.03737427840266619,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 157
    },
    {
      "epoch": 0.03761233113134559,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 158
    },
    {
      "epoch": 0.037850383860025,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 159
    },
    {
      "epoch": 0.0380884365887044,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 160
    },
    {
      "epoch": 0.0383264893173838,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 161
    },
    {
      "epoch": 0.0385645420460632,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 162
    },
    {
      "epoch": 0.038802594774742606,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 163
    },
    {
      "epoch": 0.03904064750342201,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 164
    },
    {
      "epoch": 0.03927870023210141,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3461,
      "step": 165
    },
    {
      "epoch": 0.03951675296078081,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3745,
      "step": 166
    },
    {
      "epoch": 0.039754805689460215,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 167
    },
    {
      "epoch": 0.03999285841813962,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 168
    },
    {
      "epoch": 0.04023091114681902,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 169
    },
    {
      "epoch": 0.04046896387549842,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 170
    },
    {
      "epoch": 0.040707016604177824,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 171
    },
    {
      "epoch": 0.04094506933285723,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 172
    },
    {
      "epoch": 0.04118312206153663,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 173
    },
    {
      "epoch": 0.04142117479021603,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 174
    },
    {
      "epoch": 0.041659227518895434,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.517,
      "step": 175
    },
    {
      "epoch": 0.04189728024757484,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4576,
      "step": 176
    },
    {
      "epoch": 0.04213533297625424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 177
    },
    {
      "epoch": 0.042373385704933646,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 178
    },
    {
      "epoch": 0.04261143843361304,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 179
    },
    {
      "epoch": 0.04284949116229245,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.4582,
      "step": 180
    },
    {
      "epoch": 0.04308754389097185,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 181
    },
    {
      "epoch": 0.043325596619651255,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4674,
      "step": 182
    },
    {
      "epoch": 0.04356364934833065,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 183
    },
    {
      "epoch": 0.043801702077010056,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 184
    },
    {
      "epoch": 0.04403975480568946,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 185
    },
    {
      "epoch": 0.044277807534368864,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.4665,
      "step": 186
    },
    {
      "epoch": 0.04451586026304827,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 187
    },
    {
      "epoch": 0.044753912991727665,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.38,
      "step": 188
    },
    {
      "epoch": 0.04499196572040707,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 189
    },
    {
      "epoch": 0.04523001844908647,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 190
    },
    {
      "epoch": 0.04546807117776588,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 191
    },
    {
      "epoch": 0.045706123906445274,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 192
    },
    {
      "epoch": 0.04594417663512468,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 193
    },
    {
      "epoch": 0.04618222936380408,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 194
    },
    {
      "epoch": 0.046420282092483486,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 195
    },
    {
      "epoch": 0.04665833482116289,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3741,
      "step": 196
    },
    {
      "epoch": 0.04689638754984229,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 197
    },
    {
      "epoch": 0.04713444027852169,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 198
    },
    {
      "epoch": 0.047372493007201096,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 199
    },
    {
      "epoch": 0.0476105457358805,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 200
    },
    {
      "epoch": 0.0476105457358805,
      "eval_runtime": 0.5236,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 200
    },
    {
      "epoch": 0.0478485984645599,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3726,
      "step": 201
    },
    {
      "epoch": 0.0480866511932393,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 202
    },
    {
      "epoch": 0.048324703921918705,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 203
    },
    {
      "epoch": 0.04856275665059811,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 204
    },
    {
      "epoch": 0.04880080937927751,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 205
    },
    {
      "epoch": 0.04903886210795691,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 206
    },
    {
      "epoch": 0.049276914836636314,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 207
    },
    {
      "epoch": 0.04951496756531572,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 208
    },
    {
      "epoch": 0.04975302029399512,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 209
    },
    {
      "epoch": 0.04999107302267452,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 210
    },
    {
      "epoch": 0.05022912575135392,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 211
    },
    {
      "epoch": 0.05046717848003333,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 212
    },
    {
      "epoch": 0.05070523120871273,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2628,
      "step": 213
    },
    {
      "epoch": 0.050943283937392135,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 214
    },
    {
      "epoch": 0.05118133666607153,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 215
    },
    {
      "epoch": 0.051419389394750936,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 216
    },
    {
      "epoch": 0.05165744212343034,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 217
    },
    {
      "epoch": 0.051895494852109744,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3398,
      "step": 218
    },
    {
      "epoch": 0.05213354758078915,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 219
    },
    {
      "epoch": 0.052371600309468545,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 220
    },
    {
      "epoch": 0.05260965303814795,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 221
    },
    {
      "epoch": 0.05284770576682735,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 222
    },
    {
      "epoch": 0.05308575849550676,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 223
    },
    {
      "epoch": 0.053323811224186154,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 224
    },
    {
      "epoch": 0.05356186395286556,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 225
    },
    {
      "epoch": 0.05379991668154496,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 226
    },
    {
      "epoch": 0.05403796941022437,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 227
    },
    {
      "epoch": 0.05427602213890377,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 228
    },
    {
      "epoch": 0.05451407486758317,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4007,
      "step": 229
    },
    {
      "epoch": 0.05475212759626257,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 230
    },
    {
      "epoch": 0.054990180324941976,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 231
    },
    {
      "epoch": 0.05522823305362138,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 232
    },
    {
      "epoch": 0.05546628578230078,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 233
    },
    {
      "epoch": 0.05570433851098018,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 234
    },
    {
      "epoch": 0.055942391239659585,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 235
    },
    {
      "epoch": 0.05618044396833899,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.6883,
      "step": 236
    },
    {
      "epoch": 0.05641849669701839,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 237
    },
    {
      "epoch": 0.05665654942569779,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 238
    },
    {
      "epoch": 0.056894602154377194,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 239
    },
    {
      "epoch": 0.0571326548830566,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 240
    },
    {
      "epoch": 0.057370707611736,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4106,
      "step": 241
    },
    {
      "epoch": 0.0576087603404154,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3443,
      "step": 242
    },
    {
      "epoch": 0.0578468130690948,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 243
    },
    {
      "epoch": 0.05808486579777421,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 244
    },
    {
      "epoch": 0.05832291852645361,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.307,
      "step": 245
    },
    {
      "epoch": 0.058560971255133015,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 246
    },
    {
      "epoch": 0.05879902398381241,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 247
    },
    {
      "epoch": 0.059037076712491816,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 248
    },
    {
      "epoch": 0.05927512944117122,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.344,
      "step": 249
    },
    {
      "epoch": 0.059513182169850624,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 250
    },
    {
      "epoch": 0.05975123489853002,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3924,
      "step": 251
    },
    {
      "epoch": 0.059989287627209426,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 252
    },
    {
      "epoch": 0.06022734035588883,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 253
    },
    {
      "epoch": 0.060465393084568234,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 254
    },
    {
      "epoch": 0.06070344581324764,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2846,
      "step": 255
    },
    {
      "epoch": 0.060941498541927035,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 256
    },
    {
      "epoch": 0.06117955127060644,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 257
    },
    {
      "epoch": 0.06141760399928584,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 258
    },
    {
      "epoch": 0.06165565672796525,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 259
    },
    {
      "epoch": 0.061893709456644644,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 260
    },
    {
      "epoch": 0.06213176218532405,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 261
    },
    {
      "epoch": 0.06236981491400345,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 262
    },
    {
      "epoch": 0.06260786764268285,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 263
    },
    {
      "epoch": 0.06284592037136226,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 264
    },
    {
      "epoch": 0.06308397310004166,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3726,
      "step": 265
    },
    {
      "epoch": 0.06332202582872107,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 266
    },
    {
      "epoch": 0.06356007855740047,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 267
    },
    {
      "epoch": 0.06379813128607986,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3029,
      "step": 268
    },
    {
      "epoch": 0.06403618401475927,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 269
    },
    {
      "epoch": 0.06427423674343867,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 270
    },
    {
      "epoch": 0.06451228947211807,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 271
    },
    {
      "epoch": 0.06475034220079748,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 272
    },
    {
      "epoch": 0.06498839492947688,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 273
    },
    {
      "epoch": 0.06522644765815629,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 274
    },
    {
      "epoch": 0.06546450038683568,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2859,
      "step": 275
    },
    {
      "epoch": 0.06570255311551508,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 276
    },
    {
      "epoch": 0.06594060584419449,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 277
    },
    {
      "epoch": 0.06617865857287389,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 278
    },
    {
      "epoch": 0.0664167113015533,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 279
    },
    {
      "epoch": 0.0666547640302327,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3759,
      "step": 280
    },
    {
      "epoch": 0.0668928167589121,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 281
    },
    {
      "epoch": 0.0671308694875915,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 282
    },
    {
      "epoch": 0.0673689222162709,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5777,
      "step": 283
    },
    {
      "epoch": 0.06760697494495031,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 284
    },
    {
      "epoch": 0.06784502767362971,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.6283,
      "step": 285
    },
    {
      "epoch": 0.0680830804023091,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 286
    },
    {
      "epoch": 0.06832113313098852,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 287
    },
    {
      "epoch": 0.06855918585966791,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 288
    },
    {
      "epoch": 0.06879723858834733,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 289
    },
    {
      "epoch": 0.06903529131702672,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 290
    },
    {
      "epoch": 0.06927334404570612,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3427,
      "step": 291
    },
    {
      "epoch": 0.06951139677438553,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 292
    },
    {
      "epoch": 0.06974944950306493,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4335,
      "step": 293
    },
    {
      "epoch": 0.06998750223174433,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 294
    },
    {
      "epoch": 0.07022555496042374,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 295
    },
    {
      "epoch": 0.07046360768910313,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 296
    },
    {
      "epoch": 0.07070166041778254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4186,
      "step": 297
    },
    {
      "epoch": 0.07093971314646194,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 298
    },
    {
      "epoch": 0.07117776587514134,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 299
    },
    {
      "epoch": 0.07141581860382075,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3768,
      "step": 300
    },
    {
      "epoch": 0.07141581860382075,
      "eval_runtime": 0.5248,
      "eval_samples_per_second": 1.905,
      "eval_steps_per_second": 1.905,
      "step": 300
    },
    {
      "epoch": 0.07165387133250015,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 301
    },
    {
      "epoch": 0.07189192406117956,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3193,
      "step": 302
    },
    {
      "epoch": 0.07212997678985895,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 303
    },
    {
      "epoch": 0.07236802951853835,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 304
    },
    {
      "epoch": 0.07260608224721776,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3922,
      "step": 305
    },
    {
      "epoch": 0.07284413497589716,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 306
    },
    {
      "epoch": 0.07308218770457657,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 307
    },
    {
      "epoch": 0.07332024043325597,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3817,
      "step": 308
    },
    {
      "epoch": 0.07355829316193536,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 309
    },
    {
      "epoch": 0.07379634589061478,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 310
    },
    {
      "epoch": 0.07403439861929417,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4956,
      "step": 311
    },
    {
      "epoch": 0.07427245134797357,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 312
    },
    {
      "epoch": 0.07451050407665298,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2748,
      "step": 313
    },
    {
      "epoch": 0.07474855680533238,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 314
    },
    {
      "epoch": 0.07498660953401179,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 315
    },
    {
      "epoch": 0.07522466226269119,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4339,
      "step": 316
    },
    {
      "epoch": 0.07546271499137058,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 317
    },
    {
      "epoch": 0.07570076772005,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 318
    },
    {
      "epoch": 0.07593882044872939,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 319
    },
    {
      "epoch": 0.0761768731774088,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 320
    },
    {
      "epoch": 0.0764149259060882,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 321
    },
    {
      "epoch": 0.0766529786347676,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 322
    },
    {
      "epoch": 0.07689103136344701,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 323
    },
    {
      "epoch": 0.0771290840921264,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 324
    },
    {
      "epoch": 0.07736713682080582,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 325
    },
    {
      "epoch": 0.07760518954948521,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 326
    },
    {
      "epoch": 0.07784324227816461,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4409,
      "step": 327
    },
    {
      "epoch": 0.07808129500684402,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 328
    },
    {
      "epoch": 0.07831934773552342,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 329
    },
    {
      "epoch": 0.07855740046420281,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 330
    },
    {
      "epoch": 0.07879545319288223,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 331
    },
    {
      "epoch": 0.07903350592156162,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 332
    },
    {
      "epoch": 0.07927155865024103,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 333
    },
    {
      "epoch": 0.07950961137892043,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.565,
      "step": 334
    },
    {
      "epoch": 0.07974766410759983,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3567,
      "step": 335
    },
    {
      "epoch": 0.07998571683627924,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 336
    },
    {
      "epoch": 0.08022376956495864,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 337
    },
    {
      "epoch": 0.08046182229363805,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 338
    },
    {
      "epoch": 0.08069987502231744,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 339
    },
    {
      "epoch": 0.08093792775099684,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4891,
      "step": 340
    },
    {
      "epoch": 0.08117598047967625,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 341
    },
    {
      "epoch": 0.08141403320835565,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 342
    },
    {
      "epoch": 0.08165208593703506,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3374,
      "step": 343
    },
    {
      "epoch": 0.08189013866571446,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 344
    },
    {
      "epoch": 0.08212819139439385,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 345
    },
    {
      "epoch": 0.08236624412307327,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4098,
      "step": 346
    },
    {
      "epoch": 0.08260429685175266,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 347
    },
    {
      "epoch": 0.08284234958043206,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4135,
      "step": 348
    },
    {
      "epoch": 0.08308040230911147,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3558,
      "step": 349
    },
    {
      "epoch": 0.08331845503779087,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 350
    },
    {
      "epoch": 0.08355650776647028,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 351
    },
    {
      "epoch": 0.08379456049514968,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.6038,
      "step": 352
    },
    {
      "epoch": 0.08403261322382907,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 353
    },
    {
      "epoch": 0.08427066595250848,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 354
    },
    {
      "epoch": 0.08450871868118788,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 355
    },
    {
      "epoch": 0.08474677140986729,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 356
    },
    {
      "epoch": 0.08498482413854669,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 357
    },
    {
      "epoch": 0.08522287686722609,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.398,
      "step": 358
    },
    {
      "epoch": 0.0854609295959055,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4178,
      "step": 359
    },
    {
      "epoch": 0.0856989823245849,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 360
    },
    {
      "epoch": 0.0859370350532643,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4044,
      "step": 361
    },
    {
      "epoch": 0.0861750877819437,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4008,
      "step": 362
    },
    {
      "epoch": 0.0864131405106231,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4332,
      "step": 363
    },
    {
      "epoch": 0.08665119323930251,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 364
    },
    {
      "epoch": 0.0868892459679819,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4119,
      "step": 365
    },
    {
      "epoch": 0.0871272986966613,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 366
    },
    {
      "epoch": 0.08736535142534071,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 367
    },
    {
      "epoch": 0.08760340415402011,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 368
    },
    {
      "epoch": 0.08784145688269952,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 369
    },
    {
      "epoch": 0.08807950961137892,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 370
    },
    {
      "epoch": 0.08831756234005832,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 371
    },
    {
      "epoch": 0.08855561506873773,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 372
    },
    {
      "epoch": 0.08879366779741713,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 373
    },
    {
      "epoch": 0.08903172052609654,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 374
    },
    {
      "epoch": 0.08926977325477593,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 375
    },
    {
      "epoch": 0.08950782598345533,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 376
    },
    {
      "epoch": 0.08974587871213474,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 377
    },
    {
      "epoch": 0.08998393144081414,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 378
    },
    {
      "epoch": 0.09022198416949355,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 379
    },
    {
      "epoch": 0.09046003689817295,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 380
    },
    {
      "epoch": 0.09069808962685234,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 381
    },
    {
      "epoch": 0.09093614235553175,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 382
    },
    {
      "epoch": 0.09117419508421115,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 383
    },
    {
      "epoch": 0.09141224781289055,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 384
    },
    {
      "epoch": 0.09165030054156996,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 385
    },
    {
      "epoch": 0.09188835327024936,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 386
    },
    {
      "epoch": 0.09212640599892877,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 387
    },
    {
      "epoch": 0.09236445872760816,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4212,
      "step": 388
    },
    {
      "epoch": 0.09260251145628756,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.433,
      "step": 389
    },
    {
      "epoch": 0.09284056418496697,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 390
    },
    {
      "epoch": 0.09307861691364637,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3666,
      "step": 391
    },
    {
      "epoch": 0.09331666964232578,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 392
    },
    {
      "epoch": 0.09355472237100518,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4914,
      "step": 393
    },
    {
      "epoch": 0.09379277509968457,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.431,
      "step": 394
    },
    {
      "epoch": 0.09403082782836399,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 395
    },
    {
      "epoch": 0.09426888055704338,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5069,
      "step": 396
    },
    {
      "epoch": 0.0945069332857228,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 397
    },
    {
      "epoch": 0.09474498601440219,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3375,
      "step": 398
    },
    {
      "epoch": 0.09498303874308159,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 399
    },
    {
      "epoch": 0.095221091471761,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 400
    },
    {
      "epoch": 0.095221091471761,
      "eval_runtime": 0.5279,
      "eval_samples_per_second": 1.894,
      "eval_steps_per_second": 1.894,
      "step": 400
    },
    {
      "epoch": 0.0954591442004404,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2747,
      "step": 401
    },
    {
      "epoch": 0.0956971969291198,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 402
    },
    {
      "epoch": 0.0959352496577992,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 403
    },
    {
      "epoch": 0.0961733023864786,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3616,
      "step": 404
    },
    {
      "epoch": 0.09641135511515801,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 405
    },
    {
      "epoch": 0.09664940784383741,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3396,
      "step": 406
    },
    {
      "epoch": 0.0968874605725168,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.5607,
      "step": 407
    },
    {
      "epoch": 0.09712551330119622,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 408
    },
    {
      "epoch": 0.09736356602987561,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3029,
      "step": 409
    },
    {
      "epoch": 0.09760161875855503,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3543,
      "step": 410
    },
    {
      "epoch": 0.09783967148723442,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 411
    },
    {
      "epoch": 0.09807772421591382,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 412
    },
    {
      "epoch": 0.09831577694459323,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 413
    },
    {
      "epoch": 0.09855382967327263,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3465,
      "step": 414
    },
    {
      "epoch": 0.09879188240195204,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3498,
      "step": 415
    },
    {
      "epoch": 0.09902993513063144,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2936,
      "step": 416
    },
    {
      "epoch": 0.09926798785931083,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 417
    },
    {
      "epoch": 0.09950604058799024,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 418
    },
    {
      "epoch": 0.09974409331666964,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 419
    },
    {
      "epoch": 0.09998214604534904,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 420
    },
    {
      "epoch": 0.10022019877402845,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 421
    },
    {
      "epoch": 0.10045825150270785,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3814,
      "step": 422
    },
    {
      "epoch": 0.10069630423138726,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 423
    },
    {
      "epoch": 0.10093435696006665,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 424
    },
    {
      "epoch": 0.10117240968874605,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 425
    },
    {
      "epoch": 0.10141046241742546,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 426
    },
    {
      "epoch": 0.10164851514610486,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3906,
      "step": 427
    },
    {
      "epoch": 0.10188656787478427,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2565,
      "step": 428
    },
    {
      "epoch": 0.10212462060346367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 429
    },
    {
      "epoch": 0.10236267333214306,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 430
    },
    {
      "epoch": 0.10260072606082248,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 431
    },
    {
      "epoch": 0.10283877878950187,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4363,
      "step": 432
    },
    {
      "epoch": 0.10307683151818128,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 433
    },
    {
      "epoch": 0.10331488424686068,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 434
    },
    {
      "epoch": 0.10355293697554008,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 435
    },
    {
      "epoch": 0.10379098970421949,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 436
    },
    {
      "epoch": 0.10402904243289889,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 437
    },
    {
      "epoch": 0.1042670951615783,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 438
    },
    {
      "epoch": 0.1045051478902577,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 439
    },
    {
      "epoch": 0.10474320061893709,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 440
    },
    {
      "epoch": 0.1049812533476165,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.419,
      "step": 441
    },
    {
      "epoch": 0.1052193060762959,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3513,
      "step": 442
    },
    {
      "epoch": 0.1054573588049753,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 443
    },
    {
      "epoch": 0.1056954115336547,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 444
    },
    {
      "epoch": 0.1059334642623341,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 445
    },
    {
      "epoch": 0.10617151699101351,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 446
    },
    {
      "epoch": 0.10640956971969291,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 447
    },
    {
      "epoch": 0.10664762244837231,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 448
    },
    {
      "epoch": 0.10688567517705172,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5442,
      "step": 449
    },
    {
      "epoch": 0.10712372790573112,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 450
    },
    {
      "epoch": 0.10736178063441053,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 451
    },
    {
      "epoch": 0.10759983336308993,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 452
    },
    {
      "epoch": 0.10783788609176932,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3803,
      "step": 453
    },
    {
      "epoch": 0.10807593882044873,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 454
    },
    {
      "epoch": 0.10831399154912813,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 455
    },
    {
      "epoch": 0.10855204427780754,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 456
    },
    {
      "epoch": 0.10879009700648694,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 457
    },
    {
      "epoch": 0.10902814973516634,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3813,
      "step": 458
    },
    {
      "epoch": 0.10926620246384575,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2572,
      "step": 459
    },
    {
      "epoch": 0.10950425519252514,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 460
    },
    {
      "epoch": 0.10974230792120454,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 461
    },
    {
      "epoch": 0.10998036064988395,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 462
    },
    {
      "epoch": 0.11021841337856335,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 463
    },
    {
      "epoch": 0.11045646610724276,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 464
    },
    {
      "epoch": 0.11069451883592216,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.6207,
      "step": 465
    },
    {
      "epoch": 0.11093257156460155,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.4523,
      "step": 466
    },
    {
      "epoch": 0.11117062429328096,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 467
    },
    {
      "epoch": 0.11140867702196036,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 468
    },
    {
      "epoch": 0.11164672975063977,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 469
    },
    {
      "epoch": 0.11188478247931917,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 470
    },
    {
      "epoch": 0.11212283520799857,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4058,
      "step": 471
    },
    {
      "epoch": 0.11236088793667798,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3044,
      "step": 472
    },
    {
      "epoch": 0.11259894066535737,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 473
    },
    {
      "epoch": 0.11283699339403679,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 474
    },
    {
      "epoch": 0.11307504612271618,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4106,
      "step": 475
    },
    {
      "epoch": 0.11331309885139558,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 476
    },
    {
      "epoch": 0.11355115158007499,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4216,
      "step": 477
    },
    {
      "epoch": 0.11378920430875439,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 478
    },
    {
      "epoch": 0.11402725703743379,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 479
    },
    {
      "epoch": 0.1142653097661132,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 480
    },
    {
      "epoch": 0.1145033624947926,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3397,
      "step": 481
    },
    {
      "epoch": 0.114741415223472,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 482
    },
    {
      "epoch": 0.1149794679521514,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 483
    },
    {
      "epoch": 0.1152175206808308,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 484
    },
    {
      "epoch": 0.11545557340951021,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5594,
      "step": 485
    },
    {
      "epoch": 0.1156936261381896,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 486
    },
    {
      "epoch": 0.11593167886686902,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 487
    },
    {
      "epoch": 0.11616973159554841,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.289,
      "step": 488
    },
    {
      "epoch": 0.11640778432422781,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 489
    },
    {
      "epoch": 0.11664583705290722,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.5443,
      "step": 490
    },
    {
      "epoch": 0.11688388978158662,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 491
    },
    {
      "epoch": 0.11712194251026603,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 492
    },
    {
      "epoch": 0.11735999523894543,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 493
    },
    {
      "epoch": 0.11759804796762482,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4975,
      "step": 494
    },
    {
      "epoch": 0.11783610069630424,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 495
    },
    {
      "epoch": 0.11807415342498363,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 496
    },
    {
      "epoch": 0.11831220615366303,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.478,
      "step": 497
    },
    {
      "epoch": 0.11855025888234244,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 498
    },
    {
      "epoch": 0.11878831161102184,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 499
    },
    {
      "epoch": 0.11902636433970125,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3858,
      "step": 500
    },
    {
      "epoch": 0.11902636433970125,
      "eval_runtime": 0.5253,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 500
    },
    {
      "epoch": 0.11926441706838065,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 501
    },
    {
      "epoch": 0.11950246979706004,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3745,
      "step": 502
    },
    {
      "epoch": 0.11974052252573945,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3977,
      "step": 503
    },
    {
      "epoch": 0.11997857525441885,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 504
    },
    {
      "epoch": 0.12021662798309826,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3126,
      "step": 505
    },
    {
      "epoch": 0.12045468071177766,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 506
    },
    {
      "epoch": 0.12069273344045706,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3784,
      "step": 507
    },
    {
      "epoch": 0.12093078616913647,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 508
    },
    {
      "epoch": 0.12116883889781586,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 509
    },
    {
      "epoch": 0.12140689162649528,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3485,
      "step": 510
    },
    {
      "epoch": 0.12164494435517467,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 511
    },
    {
      "epoch": 0.12188299708385407,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3076,
      "step": 512
    },
    {
      "epoch": 0.12212104981253348,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3485,
      "step": 513
    },
    {
      "epoch": 0.12235910254121288,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 514
    },
    {
      "epoch": 0.12259715526989227,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 515
    },
    {
      "epoch": 0.12283520799857169,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 516
    },
    {
      "epoch": 0.12307326072725108,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.397,
      "step": 517
    },
    {
      "epoch": 0.1233113134559305,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3851,
      "step": 518
    },
    {
      "epoch": 0.12354936618460989,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2979,
      "step": 519
    },
    {
      "epoch": 0.12378741891328929,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 520
    },
    {
      "epoch": 0.1240254716419687,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 521
    },
    {
      "epoch": 0.1242635243706481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 522
    },
    {
      "epoch": 0.1245015770993275,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 523
    },
    {
      "epoch": 0.1247396298280069,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 524
    },
    {
      "epoch": 0.1249776825566863,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 525
    },
    {
      "epoch": 0.1252157352853657,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 526
    },
    {
      "epoch": 0.1254537880140451,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 527
    },
    {
      "epoch": 0.12569184074272452,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 528
    },
    {
      "epoch": 0.1259298934714039,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 529
    },
    {
      "epoch": 0.12616794620008331,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 530
    },
    {
      "epoch": 0.12640599892876273,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 531
    },
    {
      "epoch": 0.12664405165744214,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4725,
      "step": 532
    },
    {
      "epoch": 0.12688210438612152,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.37,
      "step": 533
    },
    {
      "epoch": 0.12712015711480093,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 534
    },
    {
      "epoch": 0.12735820984348034,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 535
    },
    {
      "epoch": 0.12759626257215972,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 536
    },
    {
      "epoch": 0.12783431530083914,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 537
    },
    {
      "epoch": 0.12807236802951855,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 538
    },
    {
      "epoch": 0.12831042075819793,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 539
    },
    {
      "epoch": 0.12854847348687734,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 540
    },
    {
      "epoch": 0.12878652621555675,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4518,
      "step": 541
    },
    {
      "epoch": 0.12902457894423613,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 542
    },
    {
      "epoch": 0.12926263167291555,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 543
    },
    {
      "epoch": 0.12950068440159496,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 544
    },
    {
      "epoch": 0.12973873713027437,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.446,
      "step": 545
    },
    {
      "epoch": 0.12997678985895375,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 546
    },
    {
      "epoch": 0.13021484258763316,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 547
    },
    {
      "epoch": 0.13045289531631257,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.556,
      "step": 548
    },
    {
      "epoch": 0.13069094804499196,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 549
    },
    {
      "epoch": 0.13092900077367137,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 550
    },
    {
      "epoch": 0.13116705350235078,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 551
    },
    {
      "epoch": 0.13140510623103016,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 552
    },
    {
      "epoch": 0.13164315895970957,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4776,
      "step": 553
    },
    {
      "epoch": 0.13188121168838898,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3246,
      "step": 554
    },
    {
      "epoch": 0.1321192644170684,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 555
    },
    {
      "epoch": 0.13235731714574778,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.6123,
      "step": 556
    },
    {
      "epoch": 0.1325953698744272,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 557
    },
    {
      "epoch": 0.1328334226031066,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3573,
      "step": 558
    },
    {
      "epoch": 0.13307147533178598,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 559
    },
    {
      "epoch": 0.1333095280604654,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 560
    },
    {
      "epoch": 0.1335475807891448,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 561
    },
    {
      "epoch": 0.1337856335178242,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 562
    },
    {
      "epoch": 0.1340236862465036,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 563
    },
    {
      "epoch": 0.134261738975183,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 564
    },
    {
      "epoch": 0.1344997917038624,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4873,
      "step": 565
    },
    {
      "epoch": 0.1347378444325418,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3031,
      "step": 566
    },
    {
      "epoch": 0.13497589716122121,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 567
    },
    {
      "epoch": 0.13521394988990063,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 568
    },
    {
      "epoch": 0.13545200261858,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 569
    },
    {
      "epoch": 0.13569005534725942,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 570
    },
    {
      "epoch": 0.13592810807593883,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 571
    },
    {
      "epoch": 0.1361661608046182,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3957,
      "step": 572
    },
    {
      "epoch": 0.13640421353329762,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 573
    },
    {
      "epoch": 0.13664226626197704,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.5768,
      "step": 574
    },
    {
      "epoch": 0.13688031899065642,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 575
    },
    {
      "epoch": 0.13711837171933583,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 576
    },
    {
      "epoch": 0.13735642444801524,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 577
    },
    {
      "epoch": 0.13759447717669465,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 578
    },
    {
      "epoch": 0.13783252990537403,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 579
    },
    {
      "epoch": 0.13807058263405345,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 580
    },
    {
      "epoch": 0.13830863536273286,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 581
    },
    {
      "epoch": 0.13854668809141224,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 582
    },
    {
      "epoch": 0.13878474082009165,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 583
    },
    {
      "epoch": 0.13902279354877106,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 584
    },
    {
      "epoch": 0.13926084627745045,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.5417,
      "step": 585
    },
    {
      "epoch": 0.13949889900612986,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 586
    },
    {
      "epoch": 0.13973695173480927,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 587
    },
    {
      "epoch": 0.13997500446348865,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 588
    },
    {
      "epoch": 0.14021305719216806,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 589
    },
    {
      "epoch": 0.14045110992084747,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.8326,
      "step": 590
    },
    {
      "epoch": 0.14068916264952688,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.4876,
      "step": 591
    },
    {
      "epoch": 0.14092721537820627,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 592
    },
    {
      "epoch": 0.14116526810688568,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4964,
      "step": 593
    },
    {
      "epoch": 0.1414033208355651,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 594
    },
    {
      "epoch": 0.14164137356424447,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5155,
      "step": 595
    },
    {
      "epoch": 0.14187942629292388,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 596
    },
    {
      "epoch": 0.1421174790216033,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 597
    },
    {
      "epoch": 0.14235553175028268,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3727,
      "step": 598
    },
    {
      "epoch": 0.1425935844789621,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4455,
      "step": 599
    },
    {
      "epoch": 0.1428316372076415,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5067,
      "step": 600
    },
    {
      "epoch": 0.1428316372076415,
      "eval_runtime": 0.5405,
      "eval_samples_per_second": 1.85,
      "eval_steps_per_second": 1.85,
      "step": 600
    },
    {
      "epoch": 0.14306968993632088,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 601
    },
    {
      "epoch": 0.1433077426650003,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 602
    },
    {
      "epoch": 0.1435457953936797,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 603
    },
    {
      "epoch": 0.14378384812235911,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.403,
      "step": 604
    },
    {
      "epoch": 0.1440219008510385,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 605
    },
    {
      "epoch": 0.1442599535797179,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 606
    },
    {
      "epoch": 0.14449800630839732,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 607
    },
    {
      "epoch": 0.1447360590370767,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 608
    },
    {
      "epoch": 0.14497411176575611,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 609
    },
    {
      "epoch": 0.14521216449443552,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 610
    },
    {
      "epoch": 0.1454502172231149,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 611
    },
    {
      "epoch": 0.14568826995179432,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 612
    },
    {
      "epoch": 0.14592632268047373,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 613
    },
    {
      "epoch": 0.14616437540915314,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.6413,
      "step": 614
    },
    {
      "epoch": 0.14640242813783252,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 615
    },
    {
      "epoch": 0.14664048086651194,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3916,
      "step": 616
    },
    {
      "epoch": 0.14687853359519135,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 617
    },
    {
      "epoch": 0.14711658632387073,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 618
    },
    {
      "epoch": 0.14735463905255014,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3736,
      "step": 619
    },
    {
      "epoch": 0.14759269178122955,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 620
    },
    {
      "epoch": 0.14783074450990893,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 621
    },
    {
      "epoch": 0.14806879723858835,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 622
    },
    {
      "epoch": 0.14830684996726776,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 623
    },
    {
      "epoch": 0.14854490269594714,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4872,
      "step": 624
    },
    {
      "epoch": 0.14878295542462655,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3858,
      "step": 625
    },
    {
      "epoch": 0.14902100815330596,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 626
    },
    {
      "epoch": 0.14925906088198537,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 627
    },
    {
      "epoch": 0.14949711361066476,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 628
    },
    {
      "epoch": 0.14973516633934417,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 629
    },
    {
      "epoch": 0.14997321906802358,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 630
    },
    {
      "epoch": 0.15021127179670296,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 631
    },
    {
      "epoch": 0.15044932452538237,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3646,
      "step": 632
    },
    {
      "epoch": 0.15068737725406178,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 633
    },
    {
      "epoch": 0.15092542998274117,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 634
    },
    {
      "epoch": 0.15116348271142058,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 635
    },
    {
      "epoch": 0.1514015354401,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 636
    },
    {
      "epoch": 0.15163958816877937,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3124,
      "step": 637
    },
    {
      "epoch": 0.15187764089745878,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 638
    },
    {
      "epoch": 0.1521156936261382,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 639
    },
    {
      "epoch": 0.1523537463548176,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 640
    },
    {
      "epoch": 0.152591799083497,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 641
    },
    {
      "epoch": 0.1528298518121764,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 642
    },
    {
      "epoch": 0.1530679045408558,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 643
    },
    {
      "epoch": 0.1533059572695352,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 644
    },
    {
      "epoch": 0.1535440099982146,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.4387,
      "step": 645
    },
    {
      "epoch": 0.15378206272689401,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 646
    },
    {
      "epoch": 0.1540201154555734,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 647
    },
    {
      "epoch": 0.1542581681842528,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 648
    },
    {
      "epoch": 0.15449622091293222,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 649
    },
    {
      "epoch": 0.15473427364161163,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 650
    },
    {
      "epoch": 0.154972326370291,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 651
    },
    {
      "epoch": 0.15521037909897042,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 652
    },
    {
      "epoch": 0.15544843182764984,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4678,
      "step": 653
    },
    {
      "epoch": 0.15568648455632922,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 654
    },
    {
      "epoch": 0.15592453728500863,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 655
    },
    {
      "epoch": 0.15616259001368804,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 656
    },
    {
      "epoch": 0.15640064274236742,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 657
    },
    {
      "epoch": 0.15663869547104683,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 658
    },
    {
      "epoch": 0.15687674819972625,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 659
    },
    {
      "epoch": 0.15711480092840563,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3632,
      "step": 660
    },
    {
      "epoch": 0.15735285365708504,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4489,
      "step": 661
    },
    {
      "epoch": 0.15759090638576445,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 662
    },
    {
      "epoch": 0.15782895911444386,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3003,
      "step": 663
    },
    {
      "epoch": 0.15806701184312325,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 664
    },
    {
      "epoch": 0.15830506457180266,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3598,
      "step": 665
    },
    {
      "epoch": 0.15854311730048207,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 666
    },
    {
      "epoch": 0.15878117002916145,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 667
    },
    {
      "epoch": 0.15901922275784086,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 668
    },
    {
      "epoch": 0.15925727548652027,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 669
    },
    {
      "epoch": 0.15949532821519966,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 670
    },
    {
      "epoch": 0.15973338094387907,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 671
    },
    {
      "epoch": 0.15997143367255848,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 672
    },
    {
      "epoch": 0.16020948640123786,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 673
    },
    {
      "epoch": 0.16044753912991727,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 674
    },
    {
      "epoch": 0.16068559185859668,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 675
    },
    {
      "epoch": 0.1609236445872761,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 676
    },
    {
      "epoch": 0.16116169731595548,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 677
    },
    {
      "epoch": 0.1613997500446349,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.384,
      "step": 678
    },
    {
      "epoch": 0.1616378027733143,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.3711,
      "step": 679
    },
    {
      "epoch": 0.16187585550199368,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 680
    },
    {
      "epoch": 0.1621139082306731,
      "grad_norm": 7.3125,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 681
    },
    {
      "epoch": 0.1623519609593525,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.5216,
      "step": 682
    },
    {
      "epoch": 0.1625900136880319,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 683
    },
    {
      "epoch": 0.1628280664167113,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 684
    },
    {
      "epoch": 0.1630661191453907,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 685
    },
    {
      "epoch": 0.16330417187407012,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 686
    },
    {
      "epoch": 0.1635422246027495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 687
    },
    {
      "epoch": 0.16378027733142891,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 688
    },
    {
      "epoch": 0.16401833006010832,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.6493,
      "step": 689
    },
    {
      "epoch": 0.1642563827887877,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 690
    },
    {
      "epoch": 0.16449443551746712,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 691
    },
    {
      "epoch": 0.16473248824614653,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 692
    },
    {
      "epoch": 0.1649705409748259,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 693
    },
    {
      "epoch": 0.16520859370350532,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 694
    },
    {
      "epoch": 0.16544664643218474,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5814,
      "step": 695
    },
    {
      "epoch": 0.16568469916086412,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 696
    },
    {
      "epoch": 0.16592275188954353,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.516,
      "step": 697
    },
    {
      "epoch": 0.16616080461822294,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 698
    },
    {
      "epoch": 0.16639885734690235,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 699
    },
    {
      "epoch": 0.16663691007558173,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 700
    },
    {
      "epoch": 0.16663691007558173,
      "eval_runtime": 0.5326,
      "eval_samples_per_second": 1.878,
      "eval_steps_per_second": 1.878,
      "step": 700
    },
    {
      "epoch": 0.16687496280426115,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 701
    },
    {
      "epoch": 0.16711301553294056,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 702
    },
    {
      "epoch": 0.16735106826161994,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 703
    },
    {
      "epoch": 0.16758912099029935,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4506,
      "step": 704
    },
    {
      "epoch": 0.16782717371897876,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4488,
      "step": 705
    },
    {
      "epoch": 0.16806522644765814,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 706
    },
    {
      "epoch": 0.16830327917633756,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 707
    },
    {
      "epoch": 0.16854133190501697,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3037,
      "step": 708
    },
    {
      "epoch": 0.16877938463369635,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 709
    },
    {
      "epoch": 0.16901743736237576,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 710
    },
    {
      "epoch": 0.16925549009105517,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 711
    },
    {
      "epoch": 0.16949354281973458,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 712
    },
    {
      "epoch": 0.16973159554841397,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 713
    },
    {
      "epoch": 0.16996964827709338,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 714
    },
    {
      "epoch": 0.1702077010057728,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 715
    },
    {
      "epoch": 0.17044575373445217,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 716
    },
    {
      "epoch": 0.17068380646313158,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4158,
      "step": 717
    },
    {
      "epoch": 0.170921859191811,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 718
    },
    {
      "epoch": 0.17115991192049038,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 719
    },
    {
      "epoch": 0.1713979646491698,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 720
    },
    {
      "epoch": 0.1716360173778492,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 721
    },
    {
      "epoch": 0.1718740701065286,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 722
    },
    {
      "epoch": 0.172112122835208,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 723
    },
    {
      "epoch": 0.1723501755638874,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 724
    },
    {
      "epoch": 0.17258822829256681,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 725
    },
    {
      "epoch": 0.1728262810212462,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 726
    },
    {
      "epoch": 0.1730643337499256,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 727
    },
    {
      "epoch": 0.17330238647860502,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 728
    },
    {
      "epoch": 0.1735404392072844,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 729
    },
    {
      "epoch": 0.1737784919359638,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 730
    },
    {
      "epoch": 0.17401654466464322,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 731
    },
    {
      "epoch": 0.1742545973933226,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 732
    },
    {
      "epoch": 0.17449265012200202,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 733
    },
    {
      "epoch": 0.17473070285068143,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 734
    },
    {
      "epoch": 0.17496875557936084,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 735
    },
    {
      "epoch": 0.17520680830804022,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 736
    },
    {
      "epoch": 0.17544486103671963,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 737
    },
    {
      "epoch": 0.17568291376539905,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 738
    },
    {
      "epoch": 0.17592096649407843,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3125,
      "step": 739
    },
    {
      "epoch": 0.17615901922275784,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 740
    },
    {
      "epoch": 0.17639707195143725,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3881,
      "step": 741
    },
    {
      "epoch": 0.17663512468011663,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 742
    },
    {
      "epoch": 0.17687317740879605,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3122,
      "step": 743
    },
    {
      "epoch": 0.17711123013747546,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.4848,
      "step": 744
    },
    {
      "epoch": 0.17734928286615487,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 745
    },
    {
      "epoch": 0.17758733559483425,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 746
    },
    {
      "epoch": 0.17782538832351366,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3073,
      "step": 747
    },
    {
      "epoch": 0.17806344105219307,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 748
    },
    {
      "epoch": 0.17830149378087246,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 749
    },
    {
      "epoch": 0.17853954650955187,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 750
    },
    {
      "epoch": 0.17877759923823128,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 751
    },
    {
      "epoch": 0.17901565196691066,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3061,
      "step": 752
    },
    {
      "epoch": 0.17925370469559007,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 753
    },
    {
      "epoch": 0.17949175742426948,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 754
    },
    {
      "epoch": 0.17972981015294887,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3873,
      "step": 755
    },
    {
      "epoch": 0.17996786288162828,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 756
    },
    {
      "epoch": 0.1802059156103077,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4785,
      "step": 757
    },
    {
      "epoch": 0.1804439683389871,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 758
    },
    {
      "epoch": 0.18068202106766648,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 759
    },
    {
      "epoch": 0.1809200737963459,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 760
    },
    {
      "epoch": 0.1811581265250253,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 761
    },
    {
      "epoch": 0.1813961792537047,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 762
    },
    {
      "epoch": 0.1816342319823841,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 763
    },
    {
      "epoch": 0.1818722847110635,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 764
    },
    {
      "epoch": 0.1821103374397429,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 765
    },
    {
      "epoch": 0.1823483901684223,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 766
    },
    {
      "epoch": 0.18258644289710171,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.386,
      "step": 767
    },
    {
      "epoch": 0.1828244956257811,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4424,
      "step": 768
    },
    {
      "epoch": 0.1830625483544605,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 769
    },
    {
      "epoch": 0.18330060108313992,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4048,
      "step": 770
    },
    {
      "epoch": 0.18353865381181933,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 771
    },
    {
      "epoch": 0.1837767065404987,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 772
    },
    {
      "epoch": 0.18401475926917812,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 773
    },
    {
      "epoch": 0.18425281199785754,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 774
    },
    {
      "epoch": 0.18449086472653692,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 775
    },
    {
      "epoch": 0.18472891745521633,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 776
    },
    {
      "epoch": 0.18496697018389574,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 777
    },
    {
      "epoch": 0.18520502291257512,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 778
    },
    {
      "epoch": 0.18544307564125453,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 779
    },
    {
      "epoch": 0.18568112836993395,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 780
    },
    {
      "epoch": 0.18591918109861336,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 781
    },
    {
      "epoch": 0.18615723382729274,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 782
    },
    {
      "epoch": 0.18639528655597215,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 783
    },
    {
      "epoch": 0.18663333928465156,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 784
    },
    {
      "epoch": 0.18687139201333094,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.513,
      "step": 785
    },
    {
      "epoch": 0.18710944474201036,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 786
    },
    {
      "epoch": 0.18734749747068977,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 787
    },
    {
      "epoch": 0.18758555019936915,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 788
    },
    {
      "epoch": 0.18782360292804856,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 789
    },
    {
      "epoch": 0.18806165565672797,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 790
    },
    {
      "epoch": 0.18829970838540736,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 791
    },
    {
      "epoch": 0.18853776111408677,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 792
    },
    {
      "epoch": 0.18877581384276618,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 793
    },
    {
      "epoch": 0.1890138665714456,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 794
    },
    {
      "epoch": 0.18925191930012497,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 795
    },
    {
      "epoch": 0.18948997202880438,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3852,
      "step": 796
    },
    {
      "epoch": 0.1897280247574838,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 797
    },
    {
      "epoch": 0.18996607748616318,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 798
    },
    {
      "epoch": 0.1902041302148426,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 799
    },
    {
      "epoch": 0.190442182943522,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 800
    },
    {
      "epoch": 0.190442182943522,
      "eval_runtime": 0.5259,
      "eval_samples_per_second": 1.901,
      "eval_steps_per_second": 1.901,
      "step": 800
    },
    {
      "epoch": 0.19068023567220138,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 801
    },
    {
      "epoch": 0.1909182884008808,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3122,
      "step": 802
    },
    {
      "epoch": 0.1911563411295602,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 803
    },
    {
      "epoch": 0.1913943938582396,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.5446,
      "step": 804
    },
    {
      "epoch": 0.191632446586919,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 805
    },
    {
      "epoch": 0.1918704993155984,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3715,
      "step": 806
    },
    {
      "epoch": 0.19210855204427782,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 807
    },
    {
      "epoch": 0.1923466047729572,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 808
    },
    {
      "epoch": 0.1925846575016366,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 809
    },
    {
      "epoch": 0.19282271023031602,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 810
    },
    {
      "epoch": 0.1930607629589954,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 811
    },
    {
      "epoch": 0.19329881568767482,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 812
    },
    {
      "epoch": 0.19353686841635423,
      "grad_norm": 8.25,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 813
    },
    {
      "epoch": 0.1937749211450336,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 814
    },
    {
      "epoch": 0.19401297387371302,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 815
    },
    {
      "epoch": 0.19425102660239243,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 816
    },
    {
      "epoch": 0.19448907933107185,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3605,
      "step": 817
    },
    {
      "epoch": 0.19472713205975123,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 818
    },
    {
      "epoch": 0.19496518478843064,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 819
    },
    {
      "epoch": 0.19520323751711005,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3265,
      "step": 820
    },
    {
      "epoch": 0.19544129024578943,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 821
    },
    {
      "epoch": 0.19567934297446885,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4669,
      "step": 822
    },
    {
      "epoch": 0.19591739570314826,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 823
    },
    {
      "epoch": 0.19615544843182764,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 824
    },
    {
      "epoch": 0.19639350116050705,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 825
    },
    {
      "epoch": 0.19663155388918646,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 826
    },
    {
      "epoch": 0.19686960661786584,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 827
    },
    {
      "epoch": 0.19710765934654526,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 828
    },
    {
      "epoch": 0.19734571207522467,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 829
    },
    {
      "epoch": 0.19758376480390408,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4035,
      "step": 830
    },
    {
      "epoch": 0.19782181753258346,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 831
    },
    {
      "epoch": 0.19805987026126287,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 832
    },
    {
      "epoch": 0.19829792298994228,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 833
    },
    {
      "epoch": 0.19853597571862167,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 834
    },
    {
      "epoch": 0.19877402844730108,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2976,
      "step": 835
    },
    {
      "epoch": 0.1990120811759805,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 836
    },
    {
      "epoch": 0.19925013390465987,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 837
    },
    {
      "epoch": 0.19948818663333928,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 838
    },
    {
      "epoch": 0.1997262393620187,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 839
    },
    {
      "epoch": 0.19996429209069808,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 840
    },
    {
      "epoch": 0.2002023448193775,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 841
    },
    {
      "epoch": 0.2004403975480569,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 842
    },
    {
      "epoch": 0.2006784502767363,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 843
    },
    {
      "epoch": 0.2009165030054157,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5736,
      "step": 844
    },
    {
      "epoch": 0.2011545557340951,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3178,
      "step": 845
    },
    {
      "epoch": 0.20139260846277451,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 846
    },
    {
      "epoch": 0.2016306611914539,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 847
    },
    {
      "epoch": 0.2018687139201333,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 848
    },
    {
      "epoch": 0.20210676664881272,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 849
    },
    {
      "epoch": 0.2023448193774921,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 850
    },
    {
      "epoch": 0.2025828721061715,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 851
    },
    {
      "epoch": 0.20282092483485092,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 852
    },
    {
      "epoch": 0.20305897756353034,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 853
    },
    {
      "epoch": 0.20329703029220972,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.5393,
      "step": 854
    },
    {
      "epoch": 0.20353508302088913,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 855
    },
    {
      "epoch": 0.20377313574956854,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5101,
      "step": 856
    },
    {
      "epoch": 0.20401118847824792,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 857
    },
    {
      "epoch": 0.20424924120692733,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3979,
      "step": 858
    },
    {
      "epoch": 0.20448729393560675,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4204,
      "step": 859
    },
    {
      "epoch": 0.20472534666428613,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 860
    },
    {
      "epoch": 0.20496339939296554,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 861
    },
    {
      "epoch": 0.20520145212164495,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 862
    },
    {
      "epoch": 0.20543950485032433,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 863
    },
    {
      "epoch": 0.20567755757900374,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 864
    },
    {
      "epoch": 0.20591561030768316,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3711,
      "step": 865
    },
    {
      "epoch": 0.20615366303636257,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 866
    },
    {
      "epoch": 0.20639171576504195,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3137,
      "step": 867
    },
    {
      "epoch": 0.20662976849372136,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3378,
      "step": 868
    },
    {
      "epoch": 0.20686782122240077,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 869
    },
    {
      "epoch": 0.20710587395108015,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 870
    },
    {
      "epoch": 0.20734392667975957,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 871
    },
    {
      "epoch": 0.20758197940843898,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 872
    },
    {
      "epoch": 0.20782003213711836,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 873
    },
    {
      "epoch": 0.20805808486579777,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 874
    },
    {
      "epoch": 0.20829613759447718,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 875
    },
    {
      "epoch": 0.2085341903231566,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 876
    },
    {
      "epoch": 0.20877224305183598,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 877
    },
    {
      "epoch": 0.2090102957805154,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 878
    },
    {
      "epoch": 0.2092483485091948,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3711,
      "step": 879
    },
    {
      "epoch": 0.20948640123787418,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 880
    },
    {
      "epoch": 0.2097244539665536,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 881
    },
    {
      "epoch": 0.209962506695233,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3493,
      "step": 882
    },
    {
      "epoch": 0.2102005594239124,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 883
    },
    {
      "epoch": 0.2104386121525918,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 884
    },
    {
      "epoch": 0.2106766648812712,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 885
    },
    {
      "epoch": 0.2109147176099506,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 886
    },
    {
      "epoch": 0.21115277033863,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 887
    },
    {
      "epoch": 0.2113908230673094,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 888
    },
    {
      "epoch": 0.21162887579598882,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 889
    },
    {
      "epoch": 0.2118669285246682,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.6172,
      "step": 890
    },
    {
      "epoch": 0.21210498125334762,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 891
    },
    {
      "epoch": 0.21234303398202703,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2525,
      "step": 892
    },
    {
      "epoch": 0.2125810867107064,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 893
    },
    {
      "epoch": 0.21281913943938582,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 894
    },
    {
      "epoch": 0.21305719216806523,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 895
    },
    {
      "epoch": 0.21329524489674462,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 896
    },
    {
      "epoch": 0.21353329762542403,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 897
    },
    {
      "epoch": 0.21377135035410344,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 898
    },
    {
      "epoch": 0.21400940308278282,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 899
    },
    {
      "epoch": 0.21424745581146223,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 900
    },
    {
      "epoch": 0.21424745581146223,
      "eval_runtime": 0.5307,
      "eval_samples_per_second": 1.884,
      "eval_steps_per_second": 1.884,
      "step": 900
    },
    {
      "epoch": 0.21448550854014165,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 901
    },
    {
      "epoch": 0.21472356126882106,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 902
    },
    {
      "epoch": 0.21496161399750044,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 903
    },
    {
      "epoch": 0.21519966672617985,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 904
    },
    {
      "epoch": 0.21543771945485926,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 905
    },
    {
      "epoch": 0.21567577218353864,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 906
    },
    {
      "epoch": 0.21591382491221806,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 907
    },
    {
      "epoch": 0.21615187764089747,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 908
    },
    {
      "epoch": 0.21638993036957685,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 909
    },
    {
      "epoch": 0.21662798309825626,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 910
    },
    {
      "epoch": 0.21686603582693567,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 911
    },
    {
      "epoch": 0.21710408855561508,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 912
    },
    {
      "epoch": 0.21734214128429447,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 913
    },
    {
      "epoch": 0.21758019401297388,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 914
    },
    {
      "epoch": 0.2178182467416533,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 915
    },
    {
      "epoch": 0.21805629947033267,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4627,
      "step": 916
    },
    {
      "epoch": 0.21829435219901208,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 917
    },
    {
      "epoch": 0.2185324049276915,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 918
    },
    {
      "epoch": 0.21877045765637088,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 919
    },
    {
      "epoch": 0.2190085103850503,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 920
    },
    {
      "epoch": 0.2192465631137297,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 921
    },
    {
      "epoch": 0.21948461584240908,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 922
    },
    {
      "epoch": 0.2197226685710885,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 923
    },
    {
      "epoch": 0.2199607212997679,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 924
    },
    {
      "epoch": 0.2201987740284473,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 925
    },
    {
      "epoch": 0.2204368267571267,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 926
    },
    {
      "epoch": 0.2206748794858061,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 927
    },
    {
      "epoch": 0.22091293221448552,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3838,
      "step": 928
    },
    {
      "epoch": 0.2211509849431649,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 929
    },
    {
      "epoch": 0.2213890376718443,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 930
    },
    {
      "epoch": 0.22162709040052372,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 931
    },
    {
      "epoch": 0.2218651431292031,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 932
    },
    {
      "epoch": 0.22210319585788252,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 933
    },
    {
      "epoch": 0.22234124858656193,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 934
    },
    {
      "epoch": 0.2225793013152413,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3301,
      "step": 935
    },
    {
      "epoch": 0.22281735404392072,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 936
    },
    {
      "epoch": 0.22305540677260013,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 937
    },
    {
      "epoch": 0.22329345950127955,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 938
    },
    {
      "epoch": 0.22353151222995893,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2235,
      "step": 939
    },
    {
      "epoch": 0.22376956495863834,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 940
    },
    {
      "epoch": 0.22400761768731775,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 941
    },
    {
      "epoch": 0.22424567041599713,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 942
    },
    {
      "epoch": 0.22448372314467654,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 943
    },
    {
      "epoch": 0.22472177587335596,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 944
    },
    {
      "epoch": 0.22495982860203534,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 945
    },
    {
      "epoch": 0.22519788133071475,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4082,
      "step": 946
    },
    {
      "epoch": 0.22543593405939416,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 947
    },
    {
      "epoch": 0.22567398678807357,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 948
    },
    {
      "epoch": 0.22591203951675295,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 949
    },
    {
      "epoch": 0.22615009224543237,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 950
    },
    {
      "epoch": 0.22638814497411178,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 951
    },
    {
      "epoch": 0.22662619770279116,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 952
    },
    {
      "epoch": 0.22686425043147057,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 953
    },
    {
      "epoch": 0.22710230316014998,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4337,
      "step": 954
    },
    {
      "epoch": 0.22734035588882937,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 955
    },
    {
      "epoch": 0.22757840861750878,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 956
    },
    {
      "epoch": 0.2278164613461882,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 957
    },
    {
      "epoch": 0.22805451407486757,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4098,
      "step": 958
    },
    {
      "epoch": 0.22829256680354698,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 959
    },
    {
      "epoch": 0.2285306195322264,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 960
    },
    {
      "epoch": 0.2287686722609058,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5932,
      "step": 961
    },
    {
      "epoch": 0.2290067249895852,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 962
    },
    {
      "epoch": 0.2292447777182646,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 963
    },
    {
      "epoch": 0.229482830446944,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 964
    },
    {
      "epoch": 0.2297208831756234,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 965
    },
    {
      "epoch": 0.2299589359043028,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 966
    },
    {
      "epoch": 0.2301969886329822,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.4556,
      "step": 967
    },
    {
      "epoch": 0.2304350413616616,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 968
    },
    {
      "epoch": 0.230673094090341,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 969
    },
    {
      "epoch": 0.23091114681902042,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 970
    },
    {
      "epoch": 0.2311491995476998,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2477,
      "step": 971
    },
    {
      "epoch": 0.2313872522763792,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 972
    },
    {
      "epoch": 0.23162530500505862,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 973
    },
    {
      "epoch": 0.23186335773373803,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.6137,
      "step": 974
    },
    {
      "epoch": 0.23210141046241742,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3595,
      "step": 975
    },
    {
      "epoch": 0.23233946319109683,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 976
    },
    {
      "epoch": 0.23257751591977624,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 977
    },
    {
      "epoch": 0.23281556864845562,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.5926,
      "step": 978
    },
    {
      "epoch": 0.23305362137713503,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 979
    },
    {
      "epoch": 0.23329167410581444,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 980
    },
    {
      "epoch": 0.23352972683449383,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 981
    },
    {
      "epoch": 0.23376777956317324,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4329,
      "step": 982
    },
    {
      "epoch": 0.23400583229185265,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3597,
      "step": 983
    },
    {
      "epoch": 0.23424388502053206,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 984
    },
    {
      "epoch": 0.23448193774921144,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 985
    },
    {
      "epoch": 0.23471999047789086,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 986
    },
    {
      "epoch": 0.23495804320657027,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.5373,
      "step": 987
    },
    {
      "epoch": 0.23519609593524965,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 988
    },
    {
      "epoch": 0.23543414866392906,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 989
    },
    {
      "epoch": 0.23567220139260847,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 990
    },
    {
      "epoch": 0.23591025412128785,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 991
    },
    {
      "epoch": 0.23614830684996727,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2119,
      "step": 992
    },
    {
      "epoch": 0.23638635957864668,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 993
    },
    {
      "epoch": 0.23662441230732606,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 994
    },
    {
      "epoch": 0.23686246503600547,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 995
    },
    {
      "epoch": 0.23710051776468488,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 996
    },
    {
      "epoch": 0.2373385704933643,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 997
    },
    {
      "epoch": 0.23757662322204368,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 998
    },
    {
      "epoch": 0.2378146759507231,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 999
    },
    {
      "epoch": 0.2380527286794025,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 1000
    },
    {
      "epoch": 0.2380527286794025,
      "eval_runtime": 0.5337,
      "eval_samples_per_second": 1.874,
      "eval_steps_per_second": 1.874,
      "step": 1000
    },
    {
      "epoch": 0.23829078140808188,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 1001
    },
    {
      "epoch": 0.2385288341367613,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 1002
    },
    {
      "epoch": 0.2387668868654407,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 1003
    },
    {
      "epoch": 0.23900493959412009,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 1004
    },
    {
      "epoch": 0.2392429923227995,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3368,
      "step": 1005
    },
    {
      "epoch": 0.2394810450514789,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 1006
    },
    {
      "epoch": 0.23971909778015832,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 1007
    },
    {
      "epoch": 0.2399571505088377,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 1008
    },
    {
      "epoch": 0.2401952032375171,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 1009
    },
    {
      "epoch": 0.24043325596619652,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 1010
    },
    {
      "epoch": 0.2406713086948759,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2947,
      "step": 1011
    },
    {
      "epoch": 0.24090936142355532,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 1012
    },
    {
      "epoch": 0.24114741415223473,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2785,
      "step": 1013
    },
    {
      "epoch": 0.2413854668809141,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 1014
    },
    {
      "epoch": 0.24162351960959352,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 1015
    },
    {
      "epoch": 0.24186157233827293,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1613,
      "step": 1016
    },
    {
      "epoch": 0.24209962506695232,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 1017
    },
    {
      "epoch": 0.24233767779563173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 1018
    },
    {
      "epoch": 0.24257573052431114,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 1019
    },
    {
      "epoch": 0.24281378325299055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3161,
      "step": 1020
    },
    {
      "epoch": 0.24305183598166993,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 1021
    },
    {
      "epoch": 0.24328988871034934,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4251,
      "step": 1022
    },
    {
      "epoch": 0.24352794143902876,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 1023
    },
    {
      "epoch": 0.24376599416770814,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 1024
    },
    {
      "epoch": 0.24400404689638755,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 1025
    },
    {
      "epoch": 0.24424209962506696,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 1026
    },
    {
      "epoch": 0.24448015235374634,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 1027
    },
    {
      "epoch": 0.24471820508242575,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 1028
    },
    {
      "epoch": 0.24495625781110517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3955,
      "step": 1029
    },
    {
      "epoch": 0.24519431053978455,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 1030
    },
    {
      "epoch": 0.24543236326846396,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 1031
    },
    {
      "epoch": 0.24567041599714337,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.5393,
      "step": 1032
    },
    {
      "epoch": 0.24590846872582278,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 1033
    },
    {
      "epoch": 0.24614652145450217,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3171,
      "step": 1034
    },
    {
      "epoch": 0.24638457418318158,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 1035
    },
    {
      "epoch": 0.246622626911861,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 1036
    },
    {
      "epoch": 0.24686067964054037,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 1037
    },
    {
      "epoch": 0.24709873236921978,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 1038
    },
    {
      "epoch": 0.2473367850978992,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 1039
    },
    {
      "epoch": 0.24757483782657858,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 1040
    },
    {
      "epoch": 0.247812890555258,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 1041
    },
    {
      "epoch": 0.2480509432839374,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 1042
    },
    {
      "epoch": 0.2482889960126168,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3431,
      "step": 1043
    },
    {
      "epoch": 0.2485270487412962,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 1044
    },
    {
      "epoch": 0.2487651014699756,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 1045
    },
    {
      "epoch": 0.249003154198655,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4114,
      "step": 1046
    },
    {
      "epoch": 0.2492412069273344,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 1047
    },
    {
      "epoch": 0.2494792596560138,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 1048
    },
    {
      "epoch": 0.24971731238469322,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 1049
    },
    {
      "epoch": 0.2499553651133726,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4317,
      "step": 1050
    },
    {
      "epoch": 0.250193417842052,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 1051
    },
    {
      "epoch": 0.2504314705707314,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 1052
    },
    {
      "epoch": 0.25066952329941083,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 1053
    },
    {
      "epoch": 0.2509075760280902,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 1054
    },
    {
      "epoch": 0.2511456287567696,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 1055
    },
    {
      "epoch": 0.25138368148544904,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 1056
    },
    {
      "epoch": 0.2516217342141284,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3696,
      "step": 1057
    },
    {
      "epoch": 0.2518597869428078,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 1058
    },
    {
      "epoch": 0.25209783967148724,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3881,
      "step": 1059
    },
    {
      "epoch": 0.25233589240016663,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 1060
    },
    {
      "epoch": 0.25257394512884607,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 1061
    },
    {
      "epoch": 0.25281199785752545,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 1062
    },
    {
      "epoch": 0.25305005058620483,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3079,
      "step": 1063
    },
    {
      "epoch": 0.25328810331488427,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4283,
      "step": 1064
    },
    {
      "epoch": 0.25352615604356366,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3356,
      "step": 1065
    },
    {
      "epoch": 0.25376420877224304,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 1066
    },
    {
      "epoch": 0.2540022615009225,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 1067
    },
    {
      "epoch": 0.25424031422960186,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 1068
    },
    {
      "epoch": 0.25447836695828124,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2603,
      "step": 1069
    },
    {
      "epoch": 0.2547164196869607,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 1070
    },
    {
      "epoch": 0.25495447241564007,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 1071
    },
    {
      "epoch": 0.25519252514431945,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 1072
    },
    {
      "epoch": 0.2554305778729989,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 1073
    },
    {
      "epoch": 0.25566863060167827,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 1074
    },
    {
      "epoch": 0.25590668333035765,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3916,
      "step": 1075
    },
    {
      "epoch": 0.2561447360590371,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 1076
    },
    {
      "epoch": 0.2563827887877165,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 1077
    },
    {
      "epoch": 0.25662084151639586,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 1078
    },
    {
      "epoch": 0.2568588942450753,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.4406,
      "step": 1079
    },
    {
      "epoch": 0.2570969469737547,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 1080
    },
    {
      "epoch": 0.25733499970243406,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 1081
    },
    {
      "epoch": 0.2575730524311135,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 1082
    },
    {
      "epoch": 0.2578111051597929,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 1083
    },
    {
      "epoch": 0.25804915788847227,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 1084
    },
    {
      "epoch": 0.2582872106171517,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 1085
    },
    {
      "epoch": 0.2585252633458311,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3786,
      "step": 1086
    },
    {
      "epoch": 0.25876331607451053,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 1087
    },
    {
      "epoch": 0.2590013688031899,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2343,
      "step": 1088
    },
    {
      "epoch": 0.2592394215318693,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 1089
    },
    {
      "epoch": 0.25947747426054873,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 1090
    },
    {
      "epoch": 0.2597155269892281,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 1091
    },
    {
      "epoch": 0.2599535797179075,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 1092
    },
    {
      "epoch": 0.26019163244658694,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 1093
    },
    {
      "epoch": 0.2604296851752663,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 1094
    },
    {
      "epoch": 0.2606677379039457,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 1095
    },
    {
      "epoch": 0.26090579063262515,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3917,
      "step": 1096
    },
    {
      "epoch": 0.26114384336130453,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4257,
      "step": 1097
    },
    {
      "epoch": 0.2613818960899839,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 1098
    },
    {
      "epoch": 0.26161994881866335,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 1099
    },
    {
      "epoch": 0.26185800154734273,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 1100
    },
    {
      "epoch": 0.26185800154734273,
      "eval_runtime": 0.526,
      "eval_samples_per_second": 1.901,
      "eval_steps_per_second": 1.901,
      "step": 1100
    },
    {
      "epoch": 0.2620960542760221,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 1101
    },
    {
      "epoch": 0.26233410700470156,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.419,
      "step": 1102
    },
    {
      "epoch": 0.26257215973338094,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 1103
    },
    {
      "epoch": 0.2628102124620603,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 1104
    },
    {
      "epoch": 0.26304826519073976,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 1105
    },
    {
      "epoch": 0.26328631791941914,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 1106
    },
    {
      "epoch": 0.2635243706480985,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 1107
    },
    {
      "epoch": 0.26376242337677797,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 1108
    },
    {
      "epoch": 0.26400047610545735,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2604,
      "step": 1109
    },
    {
      "epoch": 0.2642385288341368,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 1110
    },
    {
      "epoch": 0.26447658156281617,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 1111
    },
    {
      "epoch": 0.26471463429149555,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5193,
      "step": 1112
    },
    {
      "epoch": 0.264952687020175,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3493,
      "step": 1113
    },
    {
      "epoch": 0.2651907397488544,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 1114
    },
    {
      "epoch": 0.26542879247753376,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3527,
      "step": 1115
    },
    {
      "epoch": 0.2656668452062132,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 1116
    },
    {
      "epoch": 0.2659048979348926,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5003,
      "step": 1117
    },
    {
      "epoch": 0.26614295066357196,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4713,
      "step": 1118
    },
    {
      "epoch": 0.2663810033922514,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 1119
    },
    {
      "epoch": 0.2666190561209308,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 1120
    },
    {
      "epoch": 0.26685710884961017,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 1121
    },
    {
      "epoch": 0.2670951615782896,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 1122
    },
    {
      "epoch": 0.267333214306969,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 1123
    },
    {
      "epoch": 0.2675712670356484,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 1124
    },
    {
      "epoch": 0.2678093197643278,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 1125
    },
    {
      "epoch": 0.2680473724930072,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 1126
    },
    {
      "epoch": 0.2682854252216866,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4876,
      "step": 1127
    },
    {
      "epoch": 0.268523477950366,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4002,
      "step": 1128
    },
    {
      "epoch": 0.2687615306790454,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 1129
    },
    {
      "epoch": 0.2689995834077248,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4011,
      "step": 1130
    },
    {
      "epoch": 0.2692376361364042,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3743,
      "step": 1131
    },
    {
      "epoch": 0.2694756888650836,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 1132
    },
    {
      "epoch": 0.26971374159376305,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 1133
    },
    {
      "epoch": 0.26995179432244243,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 1134
    },
    {
      "epoch": 0.2701898470511218,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 1135
    },
    {
      "epoch": 0.27042789977980125,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 1136
    },
    {
      "epoch": 0.27066595250848063,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4103,
      "step": 1137
    },
    {
      "epoch": 0.27090400523716,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 1138
    },
    {
      "epoch": 0.27114205796583946,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3907,
      "step": 1139
    },
    {
      "epoch": 0.27138011069451884,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 1140
    },
    {
      "epoch": 0.2716181634231982,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4866,
      "step": 1141
    },
    {
      "epoch": 0.27185621615187766,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 1142
    },
    {
      "epoch": 0.27209426888055704,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 1143
    },
    {
      "epoch": 0.2723323216092364,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2796,
      "step": 1144
    },
    {
      "epoch": 0.27257037433791587,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 1145
    },
    {
      "epoch": 0.27280842706659525,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 1146
    },
    {
      "epoch": 0.27304647979527463,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3437,
      "step": 1147
    },
    {
      "epoch": 0.27328453252395407,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 1148
    },
    {
      "epoch": 0.27352258525263345,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 1149
    },
    {
      "epoch": 0.27376063798131284,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 1150
    },
    {
      "epoch": 0.2739986907099923,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4043,
      "step": 1151
    },
    {
      "epoch": 0.27423674343867166,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2822,
      "step": 1152
    },
    {
      "epoch": 0.27447479616735104,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 1153
    },
    {
      "epoch": 0.2747128488960305,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 1154
    },
    {
      "epoch": 0.27495090162470986,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 1155
    },
    {
      "epoch": 0.2751889543533893,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 1156
    },
    {
      "epoch": 0.2754270070820687,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 1157
    },
    {
      "epoch": 0.27566505981074807,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 1158
    },
    {
      "epoch": 0.2759031125394275,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 1159
    },
    {
      "epoch": 0.2761411652681069,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4438,
      "step": 1160
    },
    {
      "epoch": 0.2763792179967863,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 1161
    },
    {
      "epoch": 0.2766172707254657,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 1162
    },
    {
      "epoch": 0.2768553234541451,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 1163
    },
    {
      "epoch": 0.2770933761828245,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 1164
    },
    {
      "epoch": 0.2773314289115039,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 1165
    },
    {
      "epoch": 0.2775694816401833,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 1166
    },
    {
      "epoch": 0.2778075343688627,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 1167
    },
    {
      "epoch": 0.2780455870975421,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 1168
    },
    {
      "epoch": 0.2782836398262215,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 1169
    },
    {
      "epoch": 0.2785216925549009,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3442,
      "step": 1170
    },
    {
      "epoch": 0.27875974528358033,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3965,
      "step": 1171
    },
    {
      "epoch": 0.2789977980122597,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 1172
    },
    {
      "epoch": 0.2792358507409391,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4273,
      "step": 1173
    },
    {
      "epoch": 0.27947390346961853,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 1174
    },
    {
      "epoch": 0.2797119561982979,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 1175
    },
    {
      "epoch": 0.2799500089269773,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 1176
    },
    {
      "epoch": 0.28018806165565674,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 1177
    },
    {
      "epoch": 0.2804261143843361,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 1178
    },
    {
      "epoch": 0.2806641671130155,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 1179
    },
    {
      "epoch": 0.28090221984169494,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4176,
      "step": 1180
    },
    {
      "epoch": 0.2811402725703743,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 1181
    },
    {
      "epoch": 0.28137832529905377,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 1182
    },
    {
      "epoch": 0.28161637802773315,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4963,
      "step": 1183
    },
    {
      "epoch": 0.28185443075641253,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 1184
    },
    {
      "epoch": 0.28209248348509197,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 1185
    },
    {
      "epoch": 0.28233053621377135,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 1186
    },
    {
      "epoch": 0.28256858894245074,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4869,
      "step": 1187
    },
    {
      "epoch": 0.2828066416711302,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 1188
    },
    {
      "epoch": 0.28304469439980956,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 1189
    },
    {
      "epoch": 0.28328274712848894,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 1190
    },
    {
      "epoch": 0.2835207998571684,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 1191
    },
    {
      "epoch": 0.28375885258584777,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2956,
      "step": 1192
    },
    {
      "epoch": 0.28399690531452715,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4064,
      "step": 1193
    },
    {
      "epoch": 0.2842349580432066,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 1194
    },
    {
      "epoch": 0.28447301077188597,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 1195
    },
    {
      "epoch": 0.28471106350056535,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3627,
      "step": 1196
    },
    {
      "epoch": 0.2849491162292448,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4165,
      "step": 1197
    },
    {
      "epoch": 0.2851871689579242,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4261,
      "step": 1198
    },
    {
      "epoch": 0.28542522168660356,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 1199
    },
    {
      "epoch": 0.285663274415283,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 1200
    },
    {
      "epoch": 0.285663274415283,
      "eval_runtime": 0.5422,
      "eval_samples_per_second": 1.844,
      "eval_steps_per_second": 1.844,
      "step": 1200
    },
    {
      "epoch": 0.2859013271439624,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 1201
    },
    {
      "epoch": 0.28613937987264176,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 1202
    },
    {
      "epoch": 0.2863774326013212,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 1203
    },
    {
      "epoch": 0.2866154853300006,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2937,
      "step": 1204
    },
    {
      "epoch": 0.28685353805868,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 1205
    },
    {
      "epoch": 0.2870915907873594,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 1206
    },
    {
      "epoch": 0.2873296435160388,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 1207
    },
    {
      "epoch": 0.28756769624471823,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3065,
      "step": 1208
    },
    {
      "epoch": 0.2878057489733976,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 1209
    },
    {
      "epoch": 0.288043801702077,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 1210
    },
    {
      "epoch": 0.28828185443075643,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 1211
    },
    {
      "epoch": 0.2885199071594358,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 1212
    },
    {
      "epoch": 0.2887579598881152,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 1213
    },
    {
      "epoch": 0.28899601261679464,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 1214
    },
    {
      "epoch": 0.289234065345474,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 1215
    },
    {
      "epoch": 0.2894721180741534,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 1216
    },
    {
      "epoch": 0.28971017080283284,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 1217
    },
    {
      "epoch": 0.28994822353151223,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4122,
      "step": 1218
    },
    {
      "epoch": 0.2901862762601916,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 1219
    },
    {
      "epoch": 0.29042432898887105,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 1220
    },
    {
      "epoch": 0.29066238171755043,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 1221
    },
    {
      "epoch": 0.2909004344462298,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 1222
    },
    {
      "epoch": 0.29113848717490926,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 1223
    },
    {
      "epoch": 0.29137653990358864,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 1224
    },
    {
      "epoch": 0.291614592632268,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 1225
    },
    {
      "epoch": 0.29185264536094746,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 1226
    },
    {
      "epoch": 0.29209069808962684,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 1227
    },
    {
      "epoch": 0.2923287508183063,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 1228
    },
    {
      "epoch": 0.29256680354698567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.505,
      "step": 1229
    },
    {
      "epoch": 0.29280485627566505,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 1230
    },
    {
      "epoch": 0.2930429090043445,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 1231
    },
    {
      "epoch": 0.29328096173302387,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 1232
    },
    {
      "epoch": 0.29351901446170325,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 1233
    },
    {
      "epoch": 0.2937570671903827,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 1234
    },
    {
      "epoch": 0.2939951199190621,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 1235
    },
    {
      "epoch": 0.29423317264774146,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.6663,
      "step": 1236
    },
    {
      "epoch": 0.2944712253764209,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 1237
    },
    {
      "epoch": 0.2947092781051003,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 1238
    },
    {
      "epoch": 0.29494733083377966,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 1239
    },
    {
      "epoch": 0.2951853835624591,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3231,
      "step": 1240
    },
    {
      "epoch": 0.2954234362911385,
      "grad_norm": 9.5625,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 1241
    },
    {
      "epoch": 0.29566148901981787,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 1242
    },
    {
      "epoch": 0.2958995417484973,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3603,
      "step": 1243
    },
    {
      "epoch": 0.2961375944771767,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 1244
    },
    {
      "epoch": 0.2963756472058561,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3479,
      "step": 1245
    },
    {
      "epoch": 0.2966136999345355,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 1246
    },
    {
      "epoch": 0.2968517526632149,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.394,
      "step": 1247
    },
    {
      "epoch": 0.2970898053918943,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 1248
    },
    {
      "epoch": 0.2973278581205737,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 1249
    },
    {
      "epoch": 0.2975659108492531,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 1250
    },
    {
      "epoch": 0.2978039635779325,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 1251
    },
    {
      "epoch": 0.2980420163066119,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 1252
    },
    {
      "epoch": 0.2982800690352913,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 1253
    },
    {
      "epoch": 0.29851812176397075,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 1254
    },
    {
      "epoch": 0.29875617449265013,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 1255
    },
    {
      "epoch": 0.2989942272213295,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2435,
      "step": 1256
    },
    {
      "epoch": 0.29923227995000895,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 1257
    },
    {
      "epoch": 0.29947033267868833,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 1258
    },
    {
      "epoch": 0.2997083854073677,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 1259
    },
    {
      "epoch": 0.29994643813604716,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 1260
    },
    {
      "epoch": 0.30018449086472654,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 1261
    },
    {
      "epoch": 0.3004225435934059,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 1262
    },
    {
      "epoch": 0.30066059632208536,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3424,
      "step": 1263
    },
    {
      "epoch": 0.30089864905076474,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 1264
    },
    {
      "epoch": 0.3011367017794441,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 1265
    },
    {
      "epoch": 0.30137475450812357,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 1266
    },
    {
      "epoch": 0.30161280723680295,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 1267
    },
    {
      "epoch": 0.30185085996548233,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 1268
    },
    {
      "epoch": 0.30208891269416177,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.591,
      "step": 1269
    },
    {
      "epoch": 0.30232696542284115,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3499,
      "step": 1270
    },
    {
      "epoch": 0.30256501815152054,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3946,
      "step": 1271
    },
    {
      "epoch": 0.3028030708802,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 1272
    },
    {
      "epoch": 0.30304112360887936,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 1273
    },
    {
      "epoch": 0.30327917633755874,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 1274
    },
    {
      "epoch": 0.3035172290662382,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 1275
    },
    {
      "epoch": 0.30375528179491756,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 1276
    },
    {
      "epoch": 0.303993334523597,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 1277
    },
    {
      "epoch": 0.3042313872522764,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 1278
    },
    {
      "epoch": 0.30446943998095577,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 1279
    },
    {
      "epoch": 0.3047074927096352,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 1280
    },
    {
      "epoch": 0.3049455454383146,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 1281
    },
    {
      "epoch": 0.305183598166994,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 1282
    },
    {
      "epoch": 0.3054216508956734,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3191,
      "step": 1283
    },
    {
      "epoch": 0.3056597036243528,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 1284
    },
    {
      "epoch": 0.3058977563530322,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 1285
    },
    {
      "epoch": 0.3061358090817116,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 1286
    },
    {
      "epoch": 0.306373861810391,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 1287
    },
    {
      "epoch": 0.3066119145390704,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3045,
      "step": 1288
    },
    {
      "epoch": 0.3068499672677498,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 1289
    },
    {
      "epoch": 0.3070880199964292,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 1290
    },
    {
      "epoch": 0.3073260727251086,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4778,
      "step": 1291
    },
    {
      "epoch": 0.30756412545378803,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3017,
      "step": 1292
    },
    {
      "epoch": 0.3078021781824674,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 1293
    },
    {
      "epoch": 0.3080402309111468,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2855,
      "step": 1294
    },
    {
      "epoch": 0.30827828363982623,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5925,
      "step": 1295
    },
    {
      "epoch": 0.3085163363685056,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 1296
    },
    {
      "epoch": 0.308754389097185,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 1297
    },
    {
      "epoch": 0.30899244182586444,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.447,
      "step": 1298
    },
    {
      "epoch": 0.3092304945545438,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 1299
    },
    {
      "epoch": 0.30946854728322326,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 1300
    },
    {
      "epoch": 0.30946854728322326,
      "eval_runtime": 0.5397,
      "eval_samples_per_second": 1.853,
      "eval_steps_per_second": 1.853,
      "step": 1300
    },
    {
      "epoch": 0.30970660001190264,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 1301
    },
    {
      "epoch": 0.309944652740582,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 1302
    },
    {
      "epoch": 0.31018270546926147,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 1303
    },
    {
      "epoch": 0.31042075819794085,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 1304
    },
    {
      "epoch": 0.31065881092662023,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 1305
    },
    {
      "epoch": 0.31089686365529967,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 1306
    },
    {
      "epoch": 0.31113491638397905,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.286,
      "step": 1307
    },
    {
      "epoch": 0.31137296911265844,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4244,
      "step": 1308
    },
    {
      "epoch": 0.3116110218413379,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 1309
    },
    {
      "epoch": 0.31184907457001726,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 1310
    },
    {
      "epoch": 0.31208712729869664,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2932,
      "step": 1311
    },
    {
      "epoch": 0.3123251800273761,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 1312
    },
    {
      "epoch": 0.31256323275605546,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 1313
    },
    {
      "epoch": 0.31280128548473485,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 1314
    },
    {
      "epoch": 0.3130393382134143,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 1315
    },
    {
      "epoch": 0.31327739094209367,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4147,
      "step": 1316
    },
    {
      "epoch": 0.31351544367077305,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 1317
    },
    {
      "epoch": 0.3137534963994525,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 1318
    },
    {
      "epoch": 0.3139915491281319,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 1319
    },
    {
      "epoch": 0.31422960185681126,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 1320
    },
    {
      "epoch": 0.3144676545854907,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.6064,
      "step": 1321
    },
    {
      "epoch": 0.3147057073141701,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3488,
      "step": 1322
    },
    {
      "epoch": 0.3149437600428495,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 1323
    },
    {
      "epoch": 0.3151818127715289,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 1324
    },
    {
      "epoch": 0.3154198655002083,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 1325
    },
    {
      "epoch": 0.3156579182288877,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3205,
      "step": 1326
    },
    {
      "epoch": 0.3158959709575671,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7793,
      "step": 1327
    },
    {
      "epoch": 0.3161340236862465,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3866,
      "step": 1328
    },
    {
      "epoch": 0.31637207641492593,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 1329
    },
    {
      "epoch": 0.3166101291436053,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 1330
    },
    {
      "epoch": 0.3168481818722847,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3738,
      "step": 1331
    },
    {
      "epoch": 0.31708623460096413,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 1332
    },
    {
      "epoch": 0.3173242873296435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3502,
      "step": 1333
    },
    {
      "epoch": 0.3175623400583229,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 1334
    },
    {
      "epoch": 0.31780039278700234,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 1335
    },
    {
      "epoch": 0.3180384455156817,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 1336
    },
    {
      "epoch": 0.3182764982443611,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 1337
    },
    {
      "epoch": 0.31851455097304054,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 1338
    },
    {
      "epoch": 0.3187526037017199,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.5409,
      "step": 1339
    },
    {
      "epoch": 0.3189906564303993,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 1340
    },
    {
      "epoch": 0.31922870915907875,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 1341
    },
    {
      "epoch": 0.31946676188775813,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 1342
    },
    {
      "epoch": 0.3197048146164375,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 1343
    },
    {
      "epoch": 0.31994286734511695,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 1344
    },
    {
      "epoch": 0.32018092007379634,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 1345
    },
    {
      "epoch": 0.3204189728024757,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3708,
      "step": 1346
    },
    {
      "epoch": 0.32065702553115516,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2727,
      "step": 1347
    },
    {
      "epoch": 0.32089507825983454,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.7107,
      "step": 1348
    },
    {
      "epoch": 0.321133130988514,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 1349
    },
    {
      "epoch": 0.32137118371719336,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 1350
    },
    {
      "epoch": 0.32160923644587275,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 1351
    },
    {
      "epoch": 0.3218472891745522,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 1352
    },
    {
      "epoch": 0.32208534190323157,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 1353
    },
    {
      "epoch": 0.32232339463191095,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 1354
    },
    {
      "epoch": 0.3225614473605904,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.357,
      "step": 1355
    },
    {
      "epoch": 0.3227995000892698,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 1356
    },
    {
      "epoch": 0.32303755281794916,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 1357
    },
    {
      "epoch": 0.3232756055466286,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2892,
      "step": 1358
    },
    {
      "epoch": 0.323513658275308,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3412,
      "step": 1359
    },
    {
      "epoch": 0.32375171100398736,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 1360
    },
    {
      "epoch": 0.3239897637326668,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 1361
    },
    {
      "epoch": 0.3242278164613462,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 1362
    },
    {
      "epoch": 0.32446586919002557,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3775,
      "step": 1363
    },
    {
      "epoch": 0.324703921918705,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 1364
    },
    {
      "epoch": 0.3249419746473844,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3222,
      "step": 1365
    },
    {
      "epoch": 0.3251800273760638,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 1366
    },
    {
      "epoch": 0.3254180801047432,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.462,
      "step": 1367
    },
    {
      "epoch": 0.3256561328334226,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3706,
      "step": 1368
    },
    {
      "epoch": 0.325894185562102,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 1369
    },
    {
      "epoch": 0.3261322382907814,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 1370
    },
    {
      "epoch": 0.3263702910194608,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 1371
    },
    {
      "epoch": 0.32660834374814024,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 1372
    },
    {
      "epoch": 0.3268463964768196,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 1373
    },
    {
      "epoch": 0.327084449205499,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 1374
    },
    {
      "epoch": 0.32732250193417844,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 1375
    },
    {
      "epoch": 0.32756055466285783,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 1376
    },
    {
      "epoch": 0.3277986073915372,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3431,
      "step": 1377
    },
    {
      "epoch": 0.32803666012021665,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 1378
    },
    {
      "epoch": 0.32827471284889603,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3855,
      "step": 1379
    },
    {
      "epoch": 0.3285127655775754,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 1380
    },
    {
      "epoch": 0.32875081830625486,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3394,
      "step": 1381
    },
    {
      "epoch": 0.32898887103493424,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 1382
    },
    {
      "epoch": 0.3292269237636136,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 1383
    },
    {
      "epoch": 0.32946497649229306,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 1384
    },
    {
      "epoch": 0.32970302922097244,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 1385
    },
    {
      "epoch": 0.3299410819496518,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4019,
      "step": 1386
    },
    {
      "epoch": 0.33017913467833127,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 1387
    },
    {
      "epoch": 0.33041718740701065,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 1388
    },
    {
      "epoch": 0.33065524013569003,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 1389
    },
    {
      "epoch": 0.33089329286436947,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.5202,
      "step": 1390
    },
    {
      "epoch": 0.33113134559304885,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 1391
    },
    {
      "epoch": 0.33136939832172824,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2841,
      "step": 1392
    },
    {
      "epoch": 0.3316074510504077,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 1393
    },
    {
      "epoch": 0.33184550377908706,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4114,
      "step": 1394
    },
    {
      "epoch": 0.3320835565077665,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 1395
    },
    {
      "epoch": 0.3323216092364459,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 1396
    },
    {
      "epoch": 0.33255966196512526,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2707,
      "step": 1397
    },
    {
      "epoch": 0.3327977146938047,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 1398
    },
    {
      "epoch": 0.3330357674224841,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 1399
    },
    {
      "epoch": 0.33327382015116347,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 1400
    },
    {
      "epoch": 0.33327382015116347,
      "eval_runtime": 0.5445,
      "eval_samples_per_second": 1.837,
      "eval_steps_per_second": 1.837,
      "step": 1400
    },
    {
      "epoch": 0.3335118728798429,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 1401
    },
    {
      "epoch": 0.3337499256085223,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 1402
    },
    {
      "epoch": 0.3339879783372017,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 1403
    },
    {
      "epoch": 0.3342260310658811,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.364,
      "step": 1404
    },
    {
      "epoch": 0.3344640837945605,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 1405
    },
    {
      "epoch": 0.3347021365232399,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 1406
    },
    {
      "epoch": 0.3349401892519193,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 1407
    },
    {
      "epoch": 0.3351782419805987,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 1408
    },
    {
      "epoch": 0.3354162947092781,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 1409
    },
    {
      "epoch": 0.3356543474379575,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 1410
    },
    {
      "epoch": 0.3358924001666369,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.6071,
      "step": 1411
    },
    {
      "epoch": 0.3361304528953163,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 1412
    },
    {
      "epoch": 0.33636850562399573,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 1413
    },
    {
      "epoch": 0.3366065583526751,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.496,
      "step": 1414
    },
    {
      "epoch": 0.3368446110813545,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 1415
    },
    {
      "epoch": 0.33708266381003393,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 1416
    },
    {
      "epoch": 0.3373207165387133,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 1417
    },
    {
      "epoch": 0.3375587692673927,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2783,
      "step": 1418
    },
    {
      "epoch": 0.33779682199607214,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 1419
    },
    {
      "epoch": 0.3380348747247515,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 1420
    },
    {
      "epoch": 0.33827292745343096,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4236,
      "step": 1421
    },
    {
      "epoch": 0.33851098018211034,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 1422
    },
    {
      "epoch": 0.3387490329107897,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3118,
      "step": 1423
    },
    {
      "epoch": 0.33898708563946917,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 1424
    },
    {
      "epoch": 0.33922513836814855,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 1425
    },
    {
      "epoch": 0.33946319109682793,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 1426
    },
    {
      "epoch": 0.33970124382550737,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.3919,
      "step": 1427
    },
    {
      "epoch": 0.33993929655418675,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.8469,
      "step": 1428
    },
    {
      "epoch": 0.34017734928286614,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3588,
      "step": 1429
    },
    {
      "epoch": 0.3404154020115456,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 1430
    },
    {
      "epoch": 0.34065345474022496,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.5603,
      "step": 1431
    },
    {
      "epoch": 0.34089150746890434,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 1432
    },
    {
      "epoch": 0.3411295601975838,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3922,
      "step": 1433
    },
    {
      "epoch": 0.34136761292626316,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 1434
    },
    {
      "epoch": 0.34160566565494255,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 1435
    },
    {
      "epoch": 0.341843718383622,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 1436
    },
    {
      "epoch": 0.34208177111230137,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 1437
    },
    {
      "epoch": 0.34231982384098075,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 1438
    },
    {
      "epoch": 0.3425578765696602,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 1439
    },
    {
      "epoch": 0.3427959292983396,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3177,
      "step": 1440
    },
    {
      "epoch": 0.34303398202701896,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 1441
    },
    {
      "epoch": 0.3432720347556984,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 1442
    },
    {
      "epoch": 0.3435100874843778,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 1443
    },
    {
      "epoch": 0.3437481402130572,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3404,
      "step": 1444
    },
    {
      "epoch": 0.3439861929417366,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 1445
    },
    {
      "epoch": 0.344224245670416,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 1446
    },
    {
      "epoch": 0.3444622983990954,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3124,
      "step": 1447
    },
    {
      "epoch": 0.3447003511277748,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 1448
    },
    {
      "epoch": 0.3449384038564542,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7862,
      "step": 1449
    },
    {
      "epoch": 0.34517645658513363,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4522,
      "step": 1450
    },
    {
      "epoch": 0.345414509313813,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 1451
    },
    {
      "epoch": 0.3456525620424924,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 1452
    },
    {
      "epoch": 0.34589061477117183,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 1453
    },
    {
      "epoch": 0.3461286674998512,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3358,
      "step": 1454
    },
    {
      "epoch": 0.3463667202285306,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 1455
    },
    {
      "epoch": 0.34660477295721004,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4393,
      "step": 1456
    },
    {
      "epoch": 0.3468428256858894,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2244,
      "step": 1457
    },
    {
      "epoch": 0.3470808784145688,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 1458
    },
    {
      "epoch": 0.34731893114324824,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4729,
      "step": 1459
    },
    {
      "epoch": 0.3475569838719276,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 1460
    },
    {
      "epoch": 0.347795036600607,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 1461
    },
    {
      "epoch": 0.34803308932928645,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 1462
    },
    {
      "epoch": 0.34827114205796583,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 1463
    },
    {
      "epoch": 0.3485091947866452,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 1464
    },
    {
      "epoch": 0.34874724751532465,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4474,
      "step": 1465
    },
    {
      "epoch": 0.34898530024400404,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 1466
    },
    {
      "epoch": 0.3492233529726835,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4298,
      "step": 1467
    },
    {
      "epoch": 0.34946140570136286,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4112,
      "step": 1468
    },
    {
      "epoch": 0.34969945843004224,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5737,
      "step": 1469
    },
    {
      "epoch": 0.3499375111587217,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3373,
      "step": 1470
    },
    {
      "epoch": 0.35017556388740106,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 1471
    },
    {
      "epoch": 0.35041361661608045,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4206,
      "step": 1472
    },
    {
      "epoch": 0.3506516693447599,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 1473
    },
    {
      "epoch": 0.35088972207343927,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 1474
    },
    {
      "epoch": 0.35112777480211865,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 1475
    },
    {
      "epoch": 0.3513658275307981,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4107,
      "step": 1476
    },
    {
      "epoch": 0.3516038802594775,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.6999,
      "step": 1477
    },
    {
      "epoch": 0.35184193298815686,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 1478
    },
    {
      "epoch": 0.3520799857168363,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 1479
    },
    {
      "epoch": 0.3523180384455157,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 1480
    },
    {
      "epoch": 0.35255609117419506,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 1481
    },
    {
      "epoch": 0.3527941439028745,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 1482
    },
    {
      "epoch": 0.3530321966315539,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 1483
    },
    {
      "epoch": 0.35327024936023327,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 1484
    },
    {
      "epoch": 0.3535083020889127,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 1485
    },
    {
      "epoch": 0.3537463548175921,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 1486
    },
    {
      "epoch": 0.3539844075462715,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.393,
      "step": 1487
    },
    {
      "epoch": 0.3542224602749509,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3777,
      "step": 1488
    },
    {
      "epoch": 0.3544605130036303,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 1489
    },
    {
      "epoch": 0.35469856573230973,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 1490
    },
    {
      "epoch": 0.3549366184609891,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 1491
    },
    {
      "epoch": 0.3551746711896685,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 1492
    },
    {
      "epoch": 0.35541272391834794,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 1493
    },
    {
      "epoch": 0.3556507766470273,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 1494
    },
    {
      "epoch": 0.3558888293757067,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3268,
      "step": 1495
    },
    {
      "epoch": 0.35612688210438614,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 1496
    },
    {
      "epoch": 0.3563649348330655,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4491,
      "step": 1497
    },
    {
      "epoch": 0.3566029875617449,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 1498
    },
    {
      "epoch": 0.35684104029042435,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4118,
      "step": 1499
    },
    {
      "epoch": 0.35707909301910373,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 1500
    },
    {
      "epoch": 0.35707909301910373,
      "eval_runtime": 0.5352,
      "eval_samples_per_second": 1.868,
      "eval_steps_per_second": 1.868,
      "step": 1500
    },
    {
      "epoch": 0.3573171457477831,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 1501
    },
    {
      "epoch": 0.35755519847646255,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 1502
    },
    {
      "epoch": 0.35779325120514194,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.37,
      "step": 1503
    },
    {
      "epoch": 0.3580313039338213,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 1504
    },
    {
      "epoch": 0.35826935666250076,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 1505
    },
    {
      "epoch": 0.35850740939118014,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2847,
      "step": 1506
    },
    {
      "epoch": 0.3587454621198595,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 1507
    },
    {
      "epoch": 0.35898351484853896,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 1508
    },
    {
      "epoch": 0.35922156757721835,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 1509
    },
    {
      "epoch": 0.35945962030589773,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2673,
      "step": 1510
    },
    {
      "epoch": 0.35969767303457717,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 1511
    },
    {
      "epoch": 0.35993572576325655,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 1512
    },
    {
      "epoch": 0.36017377849193594,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 1513
    },
    {
      "epoch": 0.3604118312206154,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.379,
      "step": 1514
    },
    {
      "epoch": 0.36064988394929476,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 1515
    },
    {
      "epoch": 0.3608879366779742,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 1516
    },
    {
      "epoch": 0.3611259894066536,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 1517
    },
    {
      "epoch": 0.36136404213533296,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 1518
    },
    {
      "epoch": 0.3616020948640124,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 1519
    },
    {
      "epoch": 0.3618401475926918,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 1520
    },
    {
      "epoch": 0.36207820032137117,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 1521
    },
    {
      "epoch": 0.3623162530500506,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 1522
    },
    {
      "epoch": 0.36255430577873,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 1523
    },
    {
      "epoch": 0.3627923585074094,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 1524
    },
    {
      "epoch": 0.3630304112360888,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 1525
    },
    {
      "epoch": 0.3632684639647682,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 1526
    },
    {
      "epoch": 0.3635065166934476,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 1527
    },
    {
      "epoch": 0.363744569422127,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3641,
      "step": 1528
    },
    {
      "epoch": 0.3639826221508064,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 1529
    },
    {
      "epoch": 0.3642206748794858,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 1530
    },
    {
      "epoch": 0.3644587276081652,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 1531
    },
    {
      "epoch": 0.3646967803368446,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 1532
    },
    {
      "epoch": 0.364934833065524,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 1533
    },
    {
      "epoch": 0.36517288579420343,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 1534
    },
    {
      "epoch": 0.3654109385228828,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 1535
    },
    {
      "epoch": 0.3656489912515622,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3922,
      "step": 1536
    },
    {
      "epoch": 0.36588704398024163,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 1537
    },
    {
      "epoch": 0.366125096708921,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3628,
      "step": 1538
    },
    {
      "epoch": 0.36636314943760045,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 1539
    },
    {
      "epoch": 0.36660120216627984,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 1540
    },
    {
      "epoch": 0.3668392548949592,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 1541
    },
    {
      "epoch": 0.36707730762363866,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 1542
    },
    {
      "epoch": 0.36731536035231804,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 1543
    },
    {
      "epoch": 0.3675534130809974,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 1544
    },
    {
      "epoch": 0.36779146580967687,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 1545
    },
    {
      "epoch": 0.36802951853835625,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3506,
      "step": 1546
    },
    {
      "epoch": 0.36826757126703563,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5263,
      "step": 1547
    },
    {
      "epoch": 0.36850562399571507,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 1548
    },
    {
      "epoch": 0.36874367672439445,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 1549
    },
    {
      "epoch": 0.36898172945307384,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3739,
      "step": 1550
    },
    {
      "epoch": 0.3692197821817533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 1551
    },
    {
      "epoch": 0.36945783491043266,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 1552
    },
    {
      "epoch": 0.36969588763911204,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 1553
    },
    {
      "epoch": 0.3699339403677915,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 1554
    },
    {
      "epoch": 0.37017199309647086,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3961,
      "step": 1555
    },
    {
      "epoch": 0.37041004582515025,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.667,
      "step": 1556
    },
    {
      "epoch": 0.3706480985538297,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 1557
    },
    {
      "epoch": 0.37088615128250907,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 1558
    },
    {
      "epoch": 0.37112420401118845,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3862,
      "step": 1559
    },
    {
      "epoch": 0.3713622567398679,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 1560
    },
    {
      "epoch": 0.3716003094685473,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3169,
      "step": 1561
    },
    {
      "epoch": 0.3718383621972267,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 1562
    },
    {
      "epoch": 0.3720764149259061,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4888,
      "step": 1563
    },
    {
      "epoch": 0.3723144676545855,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 1564
    },
    {
      "epoch": 0.3725525203832649,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 1565
    },
    {
      "epoch": 0.3727905731119443,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 1566
    },
    {
      "epoch": 0.3730286258406237,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 1567
    },
    {
      "epoch": 0.3732666785693031,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 1568
    },
    {
      "epoch": 0.3735047312979825,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 1569
    },
    {
      "epoch": 0.3737427840266619,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 1570
    },
    {
      "epoch": 0.37398083675534133,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.5782,
      "step": 1571
    },
    {
      "epoch": 0.3742188894840207,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 1572
    },
    {
      "epoch": 0.3744569422127001,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 1573
    },
    {
      "epoch": 0.37469499494137953,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4381,
      "step": 1574
    },
    {
      "epoch": 0.3749330476700589,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 1575
    },
    {
      "epoch": 0.3751711003987383,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 1576
    },
    {
      "epoch": 0.37540915312741774,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 1577
    },
    {
      "epoch": 0.3756472058560971,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 1578
    },
    {
      "epoch": 0.3758852585847765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 1579
    },
    {
      "epoch": 0.37612331131345594,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.375,
      "step": 1580
    },
    {
      "epoch": 0.3763613640421353,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 1581
    },
    {
      "epoch": 0.3765994167708147,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 1582
    },
    {
      "epoch": 0.37683746949949415,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 1583
    },
    {
      "epoch": 0.37707552222817353,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 1584
    },
    {
      "epoch": 0.37731357495685297,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 1585
    },
    {
      "epoch": 0.37755162768553235,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4702,
      "step": 1586
    },
    {
      "epoch": 0.37778968041421174,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 1587
    },
    {
      "epoch": 0.3780277331428912,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 1588
    },
    {
      "epoch": 0.37826578587157056,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 1589
    },
    {
      "epoch": 0.37850383860024994,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.4196,
      "step": 1590
    },
    {
      "epoch": 0.3787418913289294,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 1591
    },
    {
      "epoch": 0.37897994405760876,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3192,
      "step": 1592
    },
    {
      "epoch": 0.37921799678628815,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 1593
    },
    {
      "epoch": 0.3794560495149676,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 1594
    },
    {
      "epoch": 0.37969410224364697,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3711,
      "step": 1595
    },
    {
      "epoch": 0.37993215497232635,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3668,
      "step": 1596
    },
    {
      "epoch": 0.3801702077010058,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 1597
    },
    {
      "epoch": 0.3804082604296852,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 1598
    },
    {
      "epoch": 0.38064631315836456,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3285,
      "step": 1599
    },
    {
      "epoch": 0.380884365887044,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2726,
      "step": 1600
    },
    {
      "epoch": 0.380884365887044,
      "eval_runtime": 0.5171,
      "eval_samples_per_second": 1.934,
      "eval_steps_per_second": 1.934,
      "step": 1600
    },
    {
      "epoch": 0.3811224186157234,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 1601
    },
    {
      "epoch": 0.38136047134440276,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 1602
    },
    {
      "epoch": 0.3815985240730822,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 1603
    },
    {
      "epoch": 0.3818365768017616,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 1604
    },
    {
      "epoch": 0.38207462953044097,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 1605
    },
    {
      "epoch": 0.3823126822591204,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 1606
    },
    {
      "epoch": 0.3825507349877998,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 1607
    },
    {
      "epoch": 0.3827887877164792,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 1608
    },
    {
      "epoch": 0.3830268404451586,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.5411,
      "step": 1609
    },
    {
      "epoch": 0.383264893173838,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 1610
    },
    {
      "epoch": 0.38350294590251743,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 1611
    },
    {
      "epoch": 0.3837409986311968,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 1612
    },
    {
      "epoch": 0.3839790513598762,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 1613
    },
    {
      "epoch": 0.38421710408855564,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 1614
    },
    {
      "epoch": 0.384455156817235,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 1615
    },
    {
      "epoch": 0.3846932095459144,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 1616
    },
    {
      "epoch": 0.38493126227459384,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 1617
    },
    {
      "epoch": 0.3851693150032732,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 1618
    },
    {
      "epoch": 0.3854073677319526,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 1619
    },
    {
      "epoch": 0.38564542046063205,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4772,
      "step": 1620
    },
    {
      "epoch": 0.38588347318931143,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 1621
    },
    {
      "epoch": 0.3861215259179908,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 1622
    },
    {
      "epoch": 0.38635957864667025,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 1623
    },
    {
      "epoch": 0.38659763137534964,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3798,
      "step": 1624
    },
    {
      "epoch": 0.386835684104029,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 1625
    },
    {
      "epoch": 0.38707373683270846,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 1626
    },
    {
      "epoch": 0.38731178956138784,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 1627
    },
    {
      "epoch": 0.3875498422900672,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 1628
    },
    {
      "epoch": 0.38778789501874666,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 1629
    },
    {
      "epoch": 0.38802594774742605,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.7627,
      "step": 1630
    },
    {
      "epoch": 0.38826400047610543,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 1631
    },
    {
      "epoch": 0.38850205320478487,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3329,
      "step": 1632
    },
    {
      "epoch": 0.38874010593346425,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 1633
    },
    {
      "epoch": 0.3889781586621437,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 1634
    },
    {
      "epoch": 0.3892162113908231,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 1635
    },
    {
      "epoch": 0.38945426411950246,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 1636
    },
    {
      "epoch": 0.3896923168481819,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3785,
      "step": 1637
    },
    {
      "epoch": 0.3899303695768613,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 1638
    },
    {
      "epoch": 0.39016842230554066,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3127,
      "step": 1639
    },
    {
      "epoch": 0.3904064750342201,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 1640
    },
    {
      "epoch": 0.3906445277628995,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 1641
    },
    {
      "epoch": 0.39088258049157887,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4126,
      "step": 1642
    },
    {
      "epoch": 0.3911206332202583,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 1643
    },
    {
      "epoch": 0.3913586859489377,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5539,
      "step": 1644
    },
    {
      "epoch": 0.3915967386776171,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 1645
    },
    {
      "epoch": 0.3918347914062965,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4833,
      "step": 1646
    },
    {
      "epoch": 0.3920728441349759,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 1647
    },
    {
      "epoch": 0.3923108968636553,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 1648
    },
    {
      "epoch": 0.3925489495923347,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 1649
    },
    {
      "epoch": 0.3927870023210141,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3177,
      "step": 1650
    },
    {
      "epoch": 0.3930250550496935,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 1651
    },
    {
      "epoch": 0.3932631077783729,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 1652
    },
    {
      "epoch": 0.3935011605070523,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 1653
    },
    {
      "epoch": 0.3937392132357317,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4417,
      "step": 1654
    },
    {
      "epoch": 0.3939772659644111,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 1655
    },
    {
      "epoch": 0.3942153186930905,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5869,
      "step": 1656
    },
    {
      "epoch": 0.39445337142176995,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4651,
      "step": 1657
    },
    {
      "epoch": 0.39469142415044933,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 1658
    },
    {
      "epoch": 0.3949294768791287,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 1659
    },
    {
      "epoch": 0.39516752960780815,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 1660
    },
    {
      "epoch": 0.39540558233648754,
      "grad_norm": 2.640625,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 1661
    },
    {
      "epoch": 0.3956436350651669,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 1662
    },
    {
      "epoch": 0.39588168779384636,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 1663
    },
    {
      "epoch": 0.39611974052252574,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4236,
      "step": 1664
    },
    {
      "epoch": 0.3963577932512051,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 1665
    },
    {
      "epoch": 0.39659584597988456,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 1666
    },
    {
      "epoch": 0.39683389870856395,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3458,
      "step": 1667
    },
    {
      "epoch": 0.39707195143724333,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 1668
    },
    {
      "epoch": 0.39731000416592277,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 1669
    },
    {
      "epoch": 0.39754805689460215,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 1670
    },
    {
      "epoch": 0.39778610962328154,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 1671
    },
    {
      "epoch": 0.398024162351961,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 1672
    },
    {
      "epoch": 0.39826221508064036,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 1673
    },
    {
      "epoch": 0.39850026780931974,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4073,
      "step": 1674
    },
    {
      "epoch": 0.3987383205379992,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 1675
    },
    {
      "epoch": 0.39897637326667856,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 1676
    },
    {
      "epoch": 0.39921442599535795,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 1677
    },
    {
      "epoch": 0.3994524787240374,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 1678
    },
    {
      "epoch": 0.39969053145271677,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2804,
      "step": 1679
    },
    {
      "epoch": 0.39992858418139615,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4034,
      "step": 1680
    },
    {
      "epoch": 0.4001666369100756,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 1681
    },
    {
      "epoch": 0.400404689638755,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4585,
      "step": 1682
    },
    {
      "epoch": 0.4006427423674344,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4118,
      "step": 1683
    },
    {
      "epoch": 0.4008807950961138,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 1684
    },
    {
      "epoch": 0.4011188478247932,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 1685
    },
    {
      "epoch": 0.4013569005534726,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 1686
    },
    {
      "epoch": 0.401594953282152,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 1687
    },
    {
      "epoch": 0.4018330060108314,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 1688
    },
    {
      "epoch": 0.4020710587395108,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 1689
    },
    {
      "epoch": 0.4023091114681902,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.7912,
      "step": 1690
    },
    {
      "epoch": 0.4025471641968696,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 1691
    },
    {
      "epoch": 0.40278521692554903,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 1692
    },
    {
      "epoch": 0.4030232696542284,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4021,
      "step": 1693
    },
    {
      "epoch": 0.4032613223829078,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 1694
    },
    {
      "epoch": 0.40349937511158723,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 1695
    },
    {
      "epoch": 0.4037374278402666,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 1696
    },
    {
      "epoch": 0.403975480568946,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3807,
      "step": 1697
    },
    {
      "epoch": 0.40421353329762544,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3354,
      "step": 1698
    },
    {
      "epoch": 0.4044515860263048,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 1699
    },
    {
      "epoch": 0.4046896387549842,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 1700
    },
    {
      "epoch": 0.4046896387549842,
      "eval_runtime": 0.5075,
      "eval_samples_per_second": 1.97,
      "eval_steps_per_second": 1.97,
      "step": 1700
    },
    {
      "epoch": 0.40492769148366364,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 1701
    },
    {
      "epoch": 0.405165744212343,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 1702
    },
    {
      "epoch": 0.4054037969410224,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3323,
      "step": 1703
    },
    {
      "epoch": 0.40564184966970185,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3871,
      "step": 1704
    },
    {
      "epoch": 0.40587990239838123,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 1705
    },
    {
      "epoch": 0.40611795512706067,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 1706
    },
    {
      "epoch": 0.40635600785574005,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5743,
      "step": 1707
    },
    {
      "epoch": 0.40659406058441944,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 1708
    },
    {
      "epoch": 0.4068321133130989,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4609,
      "step": 1709
    },
    {
      "epoch": 0.40707016604177826,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 1710
    },
    {
      "epoch": 0.40730821877045764,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 1711
    },
    {
      "epoch": 0.4075462714991371,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 1712
    },
    {
      "epoch": 0.40778432422781646,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 1713
    },
    {
      "epoch": 0.40802237695649585,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 1714
    },
    {
      "epoch": 0.4082604296851753,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 1715
    },
    {
      "epoch": 0.40849848241385467,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 1716
    },
    {
      "epoch": 0.40873653514253405,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 1717
    },
    {
      "epoch": 0.4089745878712135,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 1718
    },
    {
      "epoch": 0.4092126405998929,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 1719
    },
    {
      "epoch": 0.40945069332857226,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5898,
      "step": 1720
    },
    {
      "epoch": 0.4096887460572517,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 1721
    },
    {
      "epoch": 0.4099267987859311,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 1722
    },
    {
      "epoch": 0.41016485151461046,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 1723
    },
    {
      "epoch": 0.4104029042432899,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 1724
    },
    {
      "epoch": 0.4106409569719693,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 1725
    },
    {
      "epoch": 0.41087900970064867,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 1726
    },
    {
      "epoch": 0.4111170624293281,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 1727
    },
    {
      "epoch": 0.4113551151580075,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4868,
      "step": 1728
    },
    {
      "epoch": 0.41159316788668693,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4024,
      "step": 1729
    },
    {
      "epoch": 0.4118312206153663,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 1730
    },
    {
      "epoch": 0.4120692733440457,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 1731
    },
    {
      "epoch": 0.41230732607272513,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 1732
    },
    {
      "epoch": 0.4125453788014045,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3232,
      "step": 1733
    },
    {
      "epoch": 0.4127834315300839,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 1734
    },
    {
      "epoch": 0.41302148425876334,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 1735
    },
    {
      "epoch": 0.4132595369874427,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 1736
    },
    {
      "epoch": 0.4134975897161221,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 1737
    },
    {
      "epoch": 0.41373564244480154,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3155,
      "step": 1738
    },
    {
      "epoch": 0.4139736951734809,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.7259,
      "step": 1739
    },
    {
      "epoch": 0.4142117479021603,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 1740
    },
    {
      "epoch": 0.41444980063083975,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 1741
    },
    {
      "epoch": 0.41468785335951913,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 1742
    },
    {
      "epoch": 0.4149259060881985,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 1743
    },
    {
      "epoch": 0.41516395881687795,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3097,
      "step": 1744
    },
    {
      "epoch": 0.41540201154555734,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 1745
    },
    {
      "epoch": 0.4156400642742367,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5688,
      "step": 1746
    },
    {
      "epoch": 0.41587811700291616,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 1747
    },
    {
      "epoch": 0.41611616973159554,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2432,
      "step": 1748
    },
    {
      "epoch": 0.4163542224602749,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 1749
    },
    {
      "epoch": 0.41659227518895436,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5651,
      "step": 1750
    },
    {
      "epoch": 0.41683032791763375,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 1751
    },
    {
      "epoch": 0.4170683806463132,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 1752
    },
    {
      "epoch": 0.41730643337499257,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 1753
    },
    {
      "epoch": 0.41754448610367195,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 1754
    },
    {
      "epoch": 0.4177825388323514,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 1755
    },
    {
      "epoch": 0.4180205915610308,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 1756
    },
    {
      "epoch": 0.41825864428971016,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 1757
    },
    {
      "epoch": 0.4184966970183896,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 1758
    },
    {
      "epoch": 0.418734749747069,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 1759
    },
    {
      "epoch": 0.41897280247574836,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 1760
    },
    {
      "epoch": 0.4192108552044278,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 1761
    },
    {
      "epoch": 0.4194489079331072,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 1762
    },
    {
      "epoch": 0.41968696066178657,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.6202,
      "step": 1763
    },
    {
      "epoch": 0.419925013390466,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2977,
      "step": 1764
    },
    {
      "epoch": 0.4201630661191454,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 1765
    },
    {
      "epoch": 0.4204011188478248,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 1766
    },
    {
      "epoch": 0.4206391715765042,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 1767
    },
    {
      "epoch": 0.4208772243051836,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 1768
    },
    {
      "epoch": 0.421115277033863,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 1769
    },
    {
      "epoch": 0.4213533297625424,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 1770
    },
    {
      "epoch": 0.4215913824912218,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3893,
      "step": 1771
    },
    {
      "epoch": 0.4218294352199012,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 1772
    },
    {
      "epoch": 0.4220674879485806,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4324,
      "step": 1773
    },
    {
      "epoch": 0.42230554067726,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 1774
    },
    {
      "epoch": 0.4225435934059394,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 1775
    },
    {
      "epoch": 0.4227816461346188,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3652,
      "step": 1776
    },
    {
      "epoch": 0.4230196988632982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4183,
      "step": 1777
    },
    {
      "epoch": 0.42325775159197765,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 1778
    },
    {
      "epoch": 0.42349580432065703,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 1779
    },
    {
      "epoch": 0.4237338570493364,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 1780
    },
    {
      "epoch": 0.42397190977801585,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 1781
    },
    {
      "epoch": 0.42420996250669524,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4018,
      "step": 1782
    },
    {
      "epoch": 0.4244480152353746,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2948,
      "step": 1783
    },
    {
      "epoch": 0.42468606796405406,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 1784
    },
    {
      "epoch": 0.42492412069273344,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 1785
    },
    {
      "epoch": 0.4251621734214128,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.4532,
      "step": 1786
    },
    {
      "epoch": 0.42540022615009226,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 1787
    },
    {
      "epoch": 0.42563827887877165,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.561,
      "step": 1788
    },
    {
      "epoch": 0.42587633160745103,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 1789
    },
    {
      "epoch": 0.42611438433613047,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 1790
    },
    {
      "epoch": 0.42635243706480985,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 1791
    },
    {
      "epoch": 0.42659048979348924,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 1792
    },
    {
      "epoch": 0.4268285425221687,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 1793
    },
    {
      "epoch": 0.42706659525084806,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 1794
    },
    {
      "epoch": 0.42730464797952744,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 1795
    },
    {
      "epoch": 0.4275427007082069,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4992,
      "step": 1796
    },
    {
      "epoch": 0.42778075343688626,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3696,
      "step": 1797
    },
    {
      "epoch": 0.42801880616556565,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 1798
    },
    {
      "epoch": 0.4282568588942451,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 1799
    },
    {
      "epoch": 0.42849491162292447,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 1800
    },
    {
      "epoch": 0.42849491162292447,
      "eval_runtime": 0.5214,
      "eval_samples_per_second": 1.918,
      "eval_steps_per_second": 1.918,
      "step": 1800
    },
    {
      "epoch": 0.4287329643516039,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 1801
    },
    {
      "epoch": 0.4289710170802833,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 1802
    },
    {
      "epoch": 0.4292090698089627,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 1803
    },
    {
      "epoch": 0.4294471225376421,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2896,
      "step": 1804
    },
    {
      "epoch": 0.4296851752663215,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 1805
    },
    {
      "epoch": 0.4299232279950009,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 1806
    },
    {
      "epoch": 0.4301612807236803,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 1807
    },
    {
      "epoch": 0.4303993334523597,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 1808
    },
    {
      "epoch": 0.4306373861810391,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 1809
    },
    {
      "epoch": 0.4308754389097185,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 1810
    },
    {
      "epoch": 0.4311134916383979,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5104,
      "step": 1811
    },
    {
      "epoch": 0.4313515443670773,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3174,
      "step": 1812
    },
    {
      "epoch": 0.4315895970957567,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 1813
    },
    {
      "epoch": 0.4318276498244361,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 1814
    },
    {
      "epoch": 0.4320657025531155,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3685,
      "step": 1815
    },
    {
      "epoch": 0.43230375528179493,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 1816
    },
    {
      "epoch": 0.4325418080104743,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3403,
      "step": 1817
    },
    {
      "epoch": 0.4327798607391537,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 1818
    },
    {
      "epoch": 0.43301791346783314,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 1819
    },
    {
      "epoch": 0.4332559661965125,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 1820
    },
    {
      "epoch": 0.4334940189251919,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 1821
    },
    {
      "epoch": 0.43373207165387134,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 1822
    },
    {
      "epoch": 0.4339701243825507,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 1823
    },
    {
      "epoch": 0.43420817711123016,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 1824
    },
    {
      "epoch": 0.43444622983990955,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.459,
      "step": 1825
    },
    {
      "epoch": 0.43468428256858893,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 1826
    },
    {
      "epoch": 0.43492233529726837,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 1827
    },
    {
      "epoch": 0.43516038802594775,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 1828
    },
    {
      "epoch": 0.43539844075462714,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 1829
    },
    {
      "epoch": 0.4356364934833066,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 1830
    },
    {
      "epoch": 0.43587454621198596,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 1831
    },
    {
      "epoch": 0.43611259894066534,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 1832
    },
    {
      "epoch": 0.4363506516693448,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 1833
    },
    {
      "epoch": 0.43658870439802416,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.4023,
      "step": 1834
    },
    {
      "epoch": 0.43682675712670355,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 1835
    },
    {
      "epoch": 0.437064809855383,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 1836
    },
    {
      "epoch": 0.43730286258406237,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3328,
      "step": 1837
    },
    {
      "epoch": 0.43754091531274175,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.5173,
      "step": 1838
    },
    {
      "epoch": 0.4377789680414212,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 1839
    },
    {
      "epoch": 0.4380170207701006,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 1840
    },
    {
      "epoch": 0.43825507349877996,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 1841
    },
    {
      "epoch": 0.4384931262274594,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.63,
      "step": 1842
    },
    {
      "epoch": 0.4387311789561388,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 1843
    },
    {
      "epoch": 0.43896923168481816,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3515,
      "step": 1844
    },
    {
      "epoch": 0.4392072844134976,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3174,
      "step": 1845
    },
    {
      "epoch": 0.439445337142177,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3389,
      "step": 1846
    },
    {
      "epoch": 0.43968338987085637,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 1847
    },
    {
      "epoch": 0.4399214425995358,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 1848
    },
    {
      "epoch": 0.4401594953282152,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 1849
    },
    {
      "epoch": 0.4403975480568946,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 1850
    },
    {
      "epoch": 0.440635600785574,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3671,
      "step": 1851
    },
    {
      "epoch": 0.4408736535142534,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 1852
    },
    {
      "epoch": 0.44111170624293283,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 1853
    },
    {
      "epoch": 0.4413497589716122,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 1854
    },
    {
      "epoch": 0.4415878117002916,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 1855
    },
    {
      "epoch": 0.44182586442897104,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 1856
    },
    {
      "epoch": 0.4420639171576504,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 1857
    },
    {
      "epoch": 0.4423019698863298,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 1858
    },
    {
      "epoch": 0.44254002261500924,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 1859
    },
    {
      "epoch": 0.4427780753436886,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 1860
    },
    {
      "epoch": 0.443016128072368,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 1861
    },
    {
      "epoch": 0.44325418080104745,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 1862
    },
    {
      "epoch": 0.44349223352972683,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.4491,
      "step": 1863
    },
    {
      "epoch": 0.4437302862584062,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4039,
      "step": 1864
    },
    {
      "epoch": 0.44396833898708565,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4328,
      "step": 1865
    },
    {
      "epoch": 0.44420639171576504,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 1866
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 1867
    },
    {
      "epoch": 0.44468249717312386,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2621,
      "step": 1868
    },
    {
      "epoch": 0.44492054990180324,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 1869
    },
    {
      "epoch": 0.4451586026304826,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.475,
      "step": 1870
    },
    {
      "epoch": 0.44539665535916206,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 1871
    },
    {
      "epoch": 0.44563470808784145,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4123,
      "step": 1872
    },
    {
      "epoch": 0.4458727608165209,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5651,
      "step": 1873
    },
    {
      "epoch": 0.44611081354520027,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 1874
    },
    {
      "epoch": 0.44634886627387965,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 1875
    },
    {
      "epoch": 0.4465869190025591,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2894,
      "step": 1876
    },
    {
      "epoch": 0.4468249717312385,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 1877
    },
    {
      "epoch": 0.44706302445991786,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3743,
      "step": 1878
    },
    {
      "epoch": 0.4473010771885973,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4359,
      "step": 1879
    },
    {
      "epoch": 0.4475391299172767,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.3139,
      "step": 1880
    },
    {
      "epoch": 0.44777718264595606,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 1881
    },
    {
      "epoch": 0.4480152353746355,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2591,
      "step": 1882
    },
    {
      "epoch": 0.4482532881033149,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 1883
    },
    {
      "epoch": 0.44849134083199427,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.3002,
      "step": 1884
    },
    {
      "epoch": 0.4487293935606737,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 1885
    },
    {
      "epoch": 0.4489674462893531,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 1886
    },
    {
      "epoch": 0.4492054990180325,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 1887
    },
    {
      "epoch": 0.4494435517467119,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 1888
    },
    {
      "epoch": 0.4496816044753913,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.4975,
      "step": 1889
    },
    {
      "epoch": 0.4499196572040707,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 1890
    },
    {
      "epoch": 0.4501577099327501,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 1891
    },
    {
      "epoch": 0.4503957626614295,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 1892
    },
    {
      "epoch": 0.4506338153901089,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 1893
    },
    {
      "epoch": 0.4508718681187883,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 1894
    },
    {
      "epoch": 0.4511099208474677,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 1895
    },
    {
      "epoch": 0.45134797357614714,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3984,
      "step": 1896
    },
    {
      "epoch": 0.4515860263048265,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 1897
    },
    {
      "epoch": 0.4518240790335059,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4185,
      "step": 1898
    },
    {
      "epoch": 0.45206213176218535,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 1899
    },
    {
      "epoch": 0.45230018449086473,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3117,
      "step": 1900
    },
    {
      "epoch": 0.45230018449086473,
      "eval_runtime": 0.5093,
      "eval_samples_per_second": 1.964,
      "eval_steps_per_second": 1.964,
      "step": 1900
    },
    {
      "epoch": 0.4525382372195441,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 1901
    },
    {
      "epoch": 0.45277628994822355,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 1902
    },
    {
      "epoch": 0.45301434267690294,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 1903
    },
    {
      "epoch": 0.4532523954055823,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 1904
    },
    {
      "epoch": 0.45349044813426176,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 1905
    },
    {
      "epoch": 0.45372850086294114,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 1906
    },
    {
      "epoch": 0.4539665535916205,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 1907
    },
    {
      "epoch": 0.45420460632029996,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.5499,
      "step": 1908
    },
    {
      "epoch": 0.45444265904897935,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 1909
    },
    {
      "epoch": 0.45468071177765873,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 1910
    },
    {
      "epoch": 0.45491876450633817,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 1911
    },
    {
      "epoch": 0.45515681723501755,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 1912
    },
    {
      "epoch": 0.45539486996369694,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3753,
      "step": 1913
    },
    {
      "epoch": 0.4556329226923764,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 1914
    },
    {
      "epoch": 0.45587097542105576,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 1915
    },
    {
      "epoch": 0.45610902814973514,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 1916
    },
    {
      "epoch": 0.4563470808784146,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5646,
      "step": 1917
    },
    {
      "epoch": 0.45658513360709396,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 1918
    },
    {
      "epoch": 0.4568231863357734,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 1919
    },
    {
      "epoch": 0.4570612390644528,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 1920
    },
    {
      "epoch": 0.45729929179313217,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 1921
    },
    {
      "epoch": 0.4575373445218116,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 1922
    },
    {
      "epoch": 0.457775397250491,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 1923
    },
    {
      "epoch": 0.4580134499791704,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4261,
      "step": 1924
    },
    {
      "epoch": 0.4582515027078498,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 1925
    },
    {
      "epoch": 0.4584895554365292,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 1926
    },
    {
      "epoch": 0.4587276081652086,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 1927
    },
    {
      "epoch": 0.458965660893888,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 1928
    },
    {
      "epoch": 0.4592037136225674,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 1929
    },
    {
      "epoch": 0.4594417663512468,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 1930
    },
    {
      "epoch": 0.4596798190799262,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 1931
    },
    {
      "epoch": 0.4599178718086056,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 1932
    },
    {
      "epoch": 0.460155924537285,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 1933
    },
    {
      "epoch": 0.4603939772659644,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 1934
    },
    {
      "epoch": 0.4606320299946438,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 1935
    },
    {
      "epoch": 0.4608700827233232,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 1936
    },
    {
      "epoch": 0.46110813545200263,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 1937
    },
    {
      "epoch": 0.461346188180682,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3499,
      "step": 1938
    },
    {
      "epoch": 0.4615842409093614,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 1939
    },
    {
      "epoch": 0.46182229363804084,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4441,
      "step": 1940
    },
    {
      "epoch": 0.4620603463667202,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 1941
    },
    {
      "epoch": 0.4622983990953996,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4038,
      "step": 1942
    },
    {
      "epoch": 0.46253645182407904,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 1943
    },
    {
      "epoch": 0.4627745045527584,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 1944
    },
    {
      "epoch": 0.46301255728143786,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 1945
    },
    {
      "epoch": 0.46325061001011725,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 1946
    },
    {
      "epoch": 0.46348866273879663,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 1947
    },
    {
      "epoch": 0.46372671546747607,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 1948
    },
    {
      "epoch": 0.46396476819615545,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 1949
    },
    {
      "epoch": 0.46420282092483484,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 1950
    },
    {
      "epoch": 0.4644408736535143,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 1951
    },
    {
      "epoch": 0.46467892638219366,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 1952
    },
    {
      "epoch": 0.46491697911087304,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 1953
    },
    {
      "epoch": 0.4651550318395525,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 1954
    },
    {
      "epoch": 0.46539308456823186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 1955
    },
    {
      "epoch": 0.46563113729691125,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 1956
    },
    {
      "epoch": 0.4658691900255907,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 1957
    },
    {
      "epoch": 0.46610724275427007,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 1958
    },
    {
      "epoch": 0.46634529548294945,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 1959
    },
    {
      "epoch": 0.4665833482116289,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 1960
    },
    {
      "epoch": 0.4668214009403083,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 1961
    },
    {
      "epoch": 0.46705945366898766,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 1962
    },
    {
      "epoch": 0.4672975063976671,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 1963
    },
    {
      "epoch": 0.4675355591263465,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 1964
    },
    {
      "epoch": 0.46777361185502586,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.658,
      "step": 1965
    },
    {
      "epoch": 0.4680116645837053,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 1966
    },
    {
      "epoch": 0.4682497173123847,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 1967
    },
    {
      "epoch": 0.4684877700410641,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 1968
    },
    {
      "epoch": 0.4687258227697435,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 1969
    },
    {
      "epoch": 0.4689638754984229,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 1970
    },
    {
      "epoch": 0.4692019282271023,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 1971
    },
    {
      "epoch": 0.4694399809557817,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 1972
    },
    {
      "epoch": 0.4696780336844611,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 1973
    },
    {
      "epoch": 0.46991608641314053,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4335,
      "step": 1974
    },
    {
      "epoch": 0.4701541391418199,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 1975
    },
    {
      "epoch": 0.4703921918704993,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 1976
    },
    {
      "epoch": 0.47063024459917874,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4272,
      "step": 1977
    },
    {
      "epoch": 0.4708682973278581,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3928,
      "step": 1978
    },
    {
      "epoch": 0.4711063500565375,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 1979
    },
    {
      "epoch": 0.47134440278521694,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 1980
    },
    {
      "epoch": 0.4715824555138963,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 1981
    },
    {
      "epoch": 0.4718205082425757,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 1982
    },
    {
      "epoch": 0.47205856097125515,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2689,
      "step": 1983
    },
    {
      "epoch": 0.47229661369993453,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 1984
    },
    {
      "epoch": 0.4725346664286139,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 1985
    },
    {
      "epoch": 0.47277271915729335,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 1986
    },
    {
      "epoch": 0.47301077188597274,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 1987
    },
    {
      "epoch": 0.4732488246146521,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 1988
    },
    {
      "epoch": 0.47348687734333156,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 1989
    },
    {
      "epoch": 0.47372493007201094,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 1990
    },
    {
      "epoch": 0.4739629828006904,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 1991
    },
    {
      "epoch": 0.47420103552936976,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 1992
    },
    {
      "epoch": 0.47443908825804915,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 1993
    },
    {
      "epoch": 0.4746771409867286,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 1994
    },
    {
      "epoch": 0.47491519371540797,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 1995
    },
    {
      "epoch": 0.47515324644408735,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 1996
    },
    {
      "epoch": 0.4753912991727668,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3524,
      "step": 1997
    },
    {
      "epoch": 0.4756293519014462,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 1998
    },
    {
      "epoch": 0.47586740463012556,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 1999
    },
    {
      "epoch": 0.476105457358805,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4165,
      "step": 2000
    },
    {
      "epoch": 0.476105457358805,
      "eval_runtime": 0.5351,
      "eval_samples_per_second": 1.869,
      "eval_steps_per_second": 1.869,
      "step": 2000
    },
    {
      "epoch": 0.4763435100874844,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 2001
    },
    {
      "epoch": 0.47658156281616376,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 2002
    },
    {
      "epoch": 0.4768196155448432,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 2003
    },
    {
      "epoch": 0.4770576682735226,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 2004
    },
    {
      "epoch": 0.47729572100220197,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3353,
      "step": 2005
    },
    {
      "epoch": 0.4775337737308814,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3375,
      "step": 2006
    },
    {
      "epoch": 0.4777718264595608,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2631,
      "step": 2007
    },
    {
      "epoch": 0.47800987918824017,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3142,
      "step": 2008
    },
    {
      "epoch": 0.4782479319169196,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 2009
    },
    {
      "epoch": 0.478485984645599,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2988,
      "step": 2010
    },
    {
      "epoch": 0.4787240373742784,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 2011
    },
    {
      "epoch": 0.4789620901029578,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 2012
    },
    {
      "epoch": 0.4792001428316372,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 2013
    },
    {
      "epoch": 0.47943819556031664,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 2014
    },
    {
      "epoch": 0.479676248288996,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 2015
    },
    {
      "epoch": 0.4799143010176754,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 2016
    },
    {
      "epoch": 0.48015235374635484,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3807,
      "step": 2017
    },
    {
      "epoch": 0.4803904064750342,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 2018
    },
    {
      "epoch": 0.4806284592037136,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1899,
      "step": 2019
    },
    {
      "epoch": 0.48086651193239305,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 2020
    },
    {
      "epoch": 0.48110456466107243,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 2021
    },
    {
      "epoch": 0.4813426173897518,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 2022
    },
    {
      "epoch": 0.48158067011843125,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 2023
    },
    {
      "epoch": 0.48181872284711064,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 2024
    },
    {
      "epoch": 0.48205677557579,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 2025
    },
    {
      "epoch": 0.48229482830446946,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 2026
    },
    {
      "epoch": 0.48253288103314884,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 2027
    },
    {
      "epoch": 0.4827709337618282,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 2028
    },
    {
      "epoch": 0.48300898649050766,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.5592,
      "step": 2029
    },
    {
      "epoch": 0.48324703921918705,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 2030
    },
    {
      "epoch": 0.48348509194786643,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 2031
    },
    {
      "epoch": 0.48372314467654587,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 2032
    },
    {
      "epoch": 0.48396119740522525,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 2033
    },
    {
      "epoch": 0.48419925013390464,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 2034
    },
    {
      "epoch": 0.4844373028625841,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 2035
    },
    {
      "epoch": 0.48467535559126346,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.7618,
      "step": 2036
    },
    {
      "epoch": 0.48491340831994284,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 2037
    },
    {
      "epoch": 0.4851514610486223,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3885,
      "step": 2038
    },
    {
      "epoch": 0.48538951377730166,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 2039
    },
    {
      "epoch": 0.4856275665059811,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 2040
    },
    {
      "epoch": 0.4858656192346605,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 2041
    },
    {
      "epoch": 0.48610367196333987,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 2042
    },
    {
      "epoch": 0.4863417246920193,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 2043
    },
    {
      "epoch": 0.4865797774206987,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 2044
    },
    {
      "epoch": 0.4868178301493781,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 2045
    },
    {
      "epoch": 0.4870558828780575,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 2046
    },
    {
      "epoch": 0.4872939356067369,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 2047
    },
    {
      "epoch": 0.4875319883354163,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 2048
    },
    {
      "epoch": 0.4877700410640957,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 2049
    },
    {
      "epoch": 0.4880080937927751,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.383,
      "step": 2050
    },
    {
      "epoch": 0.4882461465214545,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 2051
    },
    {
      "epoch": 0.4884841992501339,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 2052
    },
    {
      "epoch": 0.4887222519788133,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 2053
    },
    {
      "epoch": 0.4889603047074927,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 2054
    },
    {
      "epoch": 0.4891983574361721,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4065,
      "step": 2055
    },
    {
      "epoch": 0.4894364101648515,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 2056
    },
    {
      "epoch": 0.4896744628935309,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 2057
    },
    {
      "epoch": 0.48991251562221033,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 2058
    },
    {
      "epoch": 0.4901505683508897,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 2059
    },
    {
      "epoch": 0.4903886210795691,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 2060
    },
    {
      "epoch": 0.49062667380824854,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 2061
    },
    {
      "epoch": 0.4908647265369279,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 2062
    },
    {
      "epoch": 0.49110277926560736,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3729,
      "step": 2063
    },
    {
      "epoch": 0.49134083199428674,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 2064
    },
    {
      "epoch": 0.4915788847229661,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 2065
    },
    {
      "epoch": 0.49181693745164556,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 2066
    },
    {
      "epoch": 0.49205499018032495,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 2067
    },
    {
      "epoch": 0.49229304290900433,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.444,
      "step": 2068
    },
    {
      "epoch": 0.49253109563768377,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 2069
    },
    {
      "epoch": 0.49276914836636315,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 2070
    },
    {
      "epoch": 0.49300720109504254,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3767,
      "step": 2071
    },
    {
      "epoch": 0.493245253823722,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 2072
    },
    {
      "epoch": 0.49348330655240136,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 2073
    },
    {
      "epoch": 0.49372135928108074,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 2074
    },
    {
      "epoch": 0.4939594120097602,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 2075
    },
    {
      "epoch": 0.49419746473843956,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 2076
    },
    {
      "epoch": 0.49443551746711895,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 2077
    },
    {
      "epoch": 0.4946735701957984,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 2078
    },
    {
      "epoch": 0.49491162292447777,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 2079
    },
    {
      "epoch": 0.49514967565315715,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 2080
    },
    {
      "epoch": 0.4953877283818366,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4171,
      "step": 2081
    },
    {
      "epoch": 0.495625781110516,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 2082
    },
    {
      "epoch": 0.49586383383919536,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 2083
    },
    {
      "epoch": 0.4961018865678748,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4656,
      "step": 2084
    },
    {
      "epoch": 0.4963399392965542,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2943,
      "step": 2085
    },
    {
      "epoch": 0.4965779920252336,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 2086
    },
    {
      "epoch": 0.496816044753913,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 2087
    },
    {
      "epoch": 0.4970540974825924,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 2088
    },
    {
      "epoch": 0.4972921502112718,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 2089
    },
    {
      "epoch": 0.4975302029399512,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.397,
      "step": 2090
    },
    {
      "epoch": 0.4977682556686306,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 2091
    },
    {
      "epoch": 0.49800630839731,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 2092
    },
    {
      "epoch": 0.4982443611259894,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 2093
    },
    {
      "epoch": 0.4984824138546688,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 2094
    },
    {
      "epoch": 0.49872046658334823,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 2095
    },
    {
      "epoch": 0.4989585193120276,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 2096
    },
    {
      "epoch": 0.499196572040707,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 2097
    },
    {
      "epoch": 0.49943462476938644,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 2098
    },
    {
      "epoch": 0.4996726774980658,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 2099
    },
    {
      "epoch": 0.4999107302267452,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4084,
      "step": 2100
    },
    {
      "epoch": 0.4999107302267452,
      "eval_runtime": 0.5248,
      "eval_samples_per_second": 1.905,
      "eval_steps_per_second": 1.905,
      "step": 2100
    },
    {
      "epoch": 0.5001487829554246,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 2101
    },
    {
      "epoch": 0.500386835684104,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 2102
    },
    {
      "epoch": 0.5006248884127834,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 2103
    },
    {
      "epoch": 0.5008629411414628,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 2104
    },
    {
      "epoch": 0.5011009938701423,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 2105
    },
    {
      "epoch": 0.5013390465988217,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 2106
    },
    {
      "epoch": 0.501577099327501,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2881,
      "step": 2107
    },
    {
      "epoch": 0.5018151520561804,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3071,
      "step": 2108
    },
    {
      "epoch": 0.5020532047848598,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 2109
    },
    {
      "epoch": 0.5022912575135392,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 2110
    },
    {
      "epoch": 0.5025293102422187,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 2111
    },
    {
      "epoch": 0.5027673629708981,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.9623,
      "step": 2112
    },
    {
      "epoch": 0.5030054156995775,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.299,
      "step": 2113
    },
    {
      "epoch": 0.5032434684282568,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 2114
    },
    {
      "epoch": 0.5034815211569362,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 2115
    },
    {
      "epoch": 0.5037195738856156,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 2116
    },
    {
      "epoch": 0.5039576266142951,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3224,
      "step": 2117
    },
    {
      "epoch": 0.5041956793429745,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 2118
    },
    {
      "epoch": 0.5044337320716539,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.5449,
      "step": 2119
    },
    {
      "epoch": 0.5046717848003333,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 2120
    },
    {
      "epoch": 0.5049098375290126,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 2121
    },
    {
      "epoch": 0.5051478902576921,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 2122
    },
    {
      "epoch": 0.5053859429863715,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2409,
      "step": 2123
    },
    {
      "epoch": 0.5056239957150509,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 2124
    },
    {
      "epoch": 0.5058620484437303,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 2125
    },
    {
      "epoch": 0.5061001011724097,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4261,
      "step": 2126
    },
    {
      "epoch": 0.506338153901089,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4564,
      "step": 2127
    },
    {
      "epoch": 0.5065762066297685,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 2128
    },
    {
      "epoch": 0.5068142593584479,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4819,
      "step": 2129
    },
    {
      "epoch": 0.5070523120871273,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 2130
    },
    {
      "epoch": 0.5072903648158067,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 2131
    },
    {
      "epoch": 0.5075284175444861,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3675,
      "step": 2132
    },
    {
      "epoch": 0.5077664702731655,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3982,
      "step": 2133
    },
    {
      "epoch": 0.508004523001845,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.4838,
      "step": 2134
    },
    {
      "epoch": 0.5082425757305243,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 2135
    },
    {
      "epoch": 0.5084806284592037,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 2136
    },
    {
      "epoch": 0.5087186811878831,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 2137
    },
    {
      "epoch": 0.5089567339165625,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3356,
      "step": 2138
    },
    {
      "epoch": 0.5091947866452419,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 2139
    },
    {
      "epoch": 0.5094328393739214,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 2140
    },
    {
      "epoch": 0.5096708921026007,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 2141
    },
    {
      "epoch": 0.5099089448312801,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 2142
    },
    {
      "epoch": 0.5101469975599595,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.332,
      "step": 2143
    },
    {
      "epoch": 0.5103850502886389,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.272,
      "step": 2144
    },
    {
      "epoch": 0.5106231030173183,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 2145
    },
    {
      "epoch": 0.5108611557459978,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 2146
    },
    {
      "epoch": 0.5110992084746772,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 2147
    },
    {
      "epoch": 0.5113372612033565,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4757,
      "step": 2148
    },
    {
      "epoch": 0.5115753139320359,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 2149
    },
    {
      "epoch": 0.5118133666607153,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 2150
    },
    {
      "epoch": 0.5120514193893948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3127,
      "step": 2151
    },
    {
      "epoch": 0.5122894721180742,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 2152
    },
    {
      "epoch": 0.5125275248467536,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 2153
    },
    {
      "epoch": 0.512765577575433,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3606,
      "step": 2154
    },
    {
      "epoch": 0.5130036303041123,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 2155
    },
    {
      "epoch": 0.5132416830327917,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3949,
      "step": 2156
    },
    {
      "epoch": 0.5134797357614712,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 2157
    },
    {
      "epoch": 0.5137177884901506,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 2158
    },
    {
      "epoch": 0.51395584121883,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0836,
      "step": 2159
    },
    {
      "epoch": 0.5141938939475094,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 2160
    },
    {
      "epoch": 0.5144319466761887,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 2161
    },
    {
      "epoch": 0.5146699994048681,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3986,
      "step": 2162
    },
    {
      "epoch": 0.5149080521335476,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 2163
    },
    {
      "epoch": 0.515146104862227,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 2164
    },
    {
      "epoch": 0.5153841575909064,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 2165
    },
    {
      "epoch": 0.5156222103195858,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 2166
    },
    {
      "epoch": 0.5158602630482652,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 2167
    },
    {
      "epoch": 0.5160983157769445,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 2168
    },
    {
      "epoch": 0.516336368505624,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 2169
    },
    {
      "epoch": 0.5165744212343034,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3571,
      "step": 2170
    },
    {
      "epoch": 0.5168124739629828,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 2171
    },
    {
      "epoch": 0.5170505266916622,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 2172
    },
    {
      "epoch": 0.5172885794203416,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 2173
    },
    {
      "epoch": 0.5175266321490211,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 2174
    },
    {
      "epoch": 0.5177646848777004,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 2175
    },
    {
      "epoch": 0.5180027376063798,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 2176
    },
    {
      "epoch": 0.5182407903350592,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.455,
      "step": 2177
    },
    {
      "epoch": 0.5184788430637386,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 2178
    },
    {
      "epoch": 0.518716895792418,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 2179
    },
    {
      "epoch": 0.5189549485210975,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 2180
    },
    {
      "epoch": 0.5191930012497769,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 2181
    },
    {
      "epoch": 0.5194310539784562,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 2182
    },
    {
      "epoch": 0.5196691067071356,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 2183
    },
    {
      "epoch": 0.519907159435815,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 2184
    },
    {
      "epoch": 0.5201452121644944,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 2185
    },
    {
      "epoch": 0.5203832648931739,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 2186
    },
    {
      "epoch": 0.5206213176218533,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 2187
    },
    {
      "epoch": 0.5208593703505326,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3826,
      "step": 2188
    },
    {
      "epoch": 0.521097423079212,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 2189
    },
    {
      "epoch": 0.5213354758078914,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3125,
      "step": 2190
    },
    {
      "epoch": 0.5215735285365708,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 2191
    },
    {
      "epoch": 0.5218115812652503,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.6775,
      "step": 2192
    },
    {
      "epoch": 0.5220496339939297,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3728,
      "step": 2193
    },
    {
      "epoch": 0.5222876867226091,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 2194
    },
    {
      "epoch": 0.5225257394512884,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 2195
    },
    {
      "epoch": 0.5227637921799678,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 2196
    },
    {
      "epoch": 0.5230018449086473,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 2197
    },
    {
      "epoch": 0.5232398976373267,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3884,
      "step": 2198
    },
    {
      "epoch": 0.5234779503660061,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 2199
    },
    {
      "epoch": 0.5237160030946855,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 2200
    },
    {
      "epoch": 0.5237160030946855,
      "eval_runtime": 0.5254,
      "eval_samples_per_second": 1.903,
      "eval_steps_per_second": 1.903,
      "step": 2200
    },
    {
      "epoch": 0.5239540558233649,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 2201
    },
    {
      "epoch": 0.5241921085520442,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 2202
    },
    {
      "epoch": 0.5244301612807237,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 2203
    },
    {
      "epoch": 0.5246682140094031,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 2204
    },
    {
      "epoch": 0.5249062667380825,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 2205
    },
    {
      "epoch": 0.5251443194667619,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 2206
    },
    {
      "epoch": 0.5253823721954413,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4817,
      "step": 2207
    },
    {
      "epoch": 0.5256204249241206,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 2208
    },
    {
      "epoch": 0.5258584776528001,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 2209
    },
    {
      "epoch": 0.5260965303814795,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 2210
    },
    {
      "epoch": 0.5263345831101589,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4827,
      "step": 2211
    },
    {
      "epoch": 0.5265726358388383,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 2212
    },
    {
      "epoch": 0.5268106885675177,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 2213
    },
    {
      "epoch": 0.527048741296197,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 2214
    },
    {
      "epoch": 0.5272867940248765,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3519,
      "step": 2215
    },
    {
      "epoch": 0.5275248467535559,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 2216
    },
    {
      "epoch": 0.5277628994822353,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 2217
    },
    {
      "epoch": 0.5280009522109147,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 2218
    },
    {
      "epoch": 0.5282390049395941,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 2219
    },
    {
      "epoch": 0.5284770576682736,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 2220
    },
    {
      "epoch": 0.528715110396953,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 2221
    },
    {
      "epoch": 0.5289531631256323,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 2222
    },
    {
      "epoch": 0.5291912158543117,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4922,
      "step": 2223
    },
    {
      "epoch": 0.5294292685829911,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 2224
    },
    {
      "epoch": 0.5296673213116705,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 2225
    },
    {
      "epoch": 0.52990537404035,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 2226
    },
    {
      "epoch": 0.5301434267690294,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4267,
      "step": 2227
    },
    {
      "epoch": 0.5303814794977088,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 2228
    },
    {
      "epoch": 0.5306195322263881,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 2229
    },
    {
      "epoch": 0.5308575849550675,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3876,
      "step": 2230
    },
    {
      "epoch": 0.5310956376837469,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.6383,
      "step": 2231
    },
    {
      "epoch": 0.5313336904124264,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 2232
    },
    {
      "epoch": 0.5315717431411058,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 2233
    },
    {
      "epoch": 0.5318097958697852,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 2234
    },
    {
      "epoch": 0.5320478485984645,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 2235
    },
    {
      "epoch": 0.5322859013271439,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1966,
      "step": 2236
    },
    {
      "epoch": 0.5325239540558233,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4,
      "step": 2237
    },
    {
      "epoch": 0.5327620067845028,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5099,
      "step": 2238
    },
    {
      "epoch": 0.5330000595131822,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 2239
    },
    {
      "epoch": 0.5332381122418616,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.5764,
      "step": 2240
    },
    {
      "epoch": 0.533476164970541,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 2241
    },
    {
      "epoch": 0.5337142176992203,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8508,
      "step": 2242
    },
    {
      "epoch": 0.5339522704278998,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 2243
    },
    {
      "epoch": 0.5341903231565792,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 2244
    },
    {
      "epoch": 0.5344283758852586,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4352,
      "step": 2245
    },
    {
      "epoch": 0.534666428613938,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 2246
    },
    {
      "epoch": 0.5349044813426174,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 2247
    },
    {
      "epoch": 0.5351425340712967,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3191,
      "step": 2248
    },
    {
      "epoch": 0.5353805867999762,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.389,
      "step": 2249
    },
    {
      "epoch": 0.5356186395286556,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 2250
    },
    {
      "epoch": 0.535856692257335,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 2251
    },
    {
      "epoch": 0.5360947449860144,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 2252
    },
    {
      "epoch": 0.5363327977146938,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 2253
    },
    {
      "epoch": 0.5365708504433732,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4505,
      "step": 2254
    },
    {
      "epoch": 0.5368089031720527,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 2255
    },
    {
      "epoch": 0.537046955900732,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3044,
      "step": 2256
    },
    {
      "epoch": 0.5372850086294114,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 2257
    },
    {
      "epoch": 0.5375230613580908,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4512,
      "step": 2258
    },
    {
      "epoch": 0.5377611140867702,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 2259
    },
    {
      "epoch": 0.5379991668154496,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 2260
    },
    {
      "epoch": 0.5382372195441291,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 2261
    },
    {
      "epoch": 0.5384752722728084,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.581,
      "step": 2262
    },
    {
      "epoch": 0.5387133250014878,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 2263
    },
    {
      "epoch": 0.5389513777301672,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 2264
    },
    {
      "epoch": 0.5391894304588466,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 2265
    },
    {
      "epoch": 0.5394274831875261,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3794,
      "step": 2266
    },
    {
      "epoch": 0.5396655359162055,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 2267
    },
    {
      "epoch": 0.5399035886448849,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3966,
      "step": 2268
    },
    {
      "epoch": 0.5401416413735642,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 2269
    },
    {
      "epoch": 0.5403796941022436,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 2270
    },
    {
      "epoch": 0.540617746830923,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4183,
      "step": 2271
    },
    {
      "epoch": 0.5408557995596025,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 2272
    },
    {
      "epoch": 0.5410938522882819,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 2273
    },
    {
      "epoch": 0.5413319050169613,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 2274
    },
    {
      "epoch": 0.5415699577456407,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.6739,
      "step": 2275
    },
    {
      "epoch": 0.54180801047432,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 2276
    },
    {
      "epoch": 0.5420460632029994,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 2277
    },
    {
      "epoch": 0.5422841159316789,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4572,
      "step": 2278
    },
    {
      "epoch": 0.5425221686603583,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.5159,
      "step": 2279
    },
    {
      "epoch": 0.5427602213890377,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 2280
    },
    {
      "epoch": 0.5429982741177171,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3484,
      "step": 2281
    },
    {
      "epoch": 0.5432363268463964,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 2282
    },
    {
      "epoch": 0.5434743795750758,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 2283
    },
    {
      "epoch": 0.5437124323037553,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 2284
    },
    {
      "epoch": 0.5439504850324347,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 2285
    },
    {
      "epoch": 0.5441885377611141,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 2286
    },
    {
      "epoch": 0.5444265904897935,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4151,
      "step": 2287
    },
    {
      "epoch": 0.5446646432184729,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 2288
    },
    {
      "epoch": 0.5449026959471523,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 2289
    },
    {
      "epoch": 0.5451407486758317,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 2290
    },
    {
      "epoch": 0.5453788014045111,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 2291
    },
    {
      "epoch": 0.5456168541331905,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 2292
    },
    {
      "epoch": 0.5458549068618699,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 2293
    },
    {
      "epoch": 0.5460929595905493,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 2294
    },
    {
      "epoch": 0.5463310123192288,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 2295
    },
    {
      "epoch": 0.5465690650479081,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 2296
    },
    {
      "epoch": 0.5468071177765875,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 2297
    },
    {
      "epoch": 0.5470451705052669,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 2298
    },
    {
      "epoch": 0.5472832232339463,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 2299
    },
    {
      "epoch": 0.5475212759626257,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 2300
    },
    {
      "epoch": 0.5475212759626257,
      "eval_runtime": 0.5097,
      "eval_samples_per_second": 1.962,
      "eval_steps_per_second": 1.962,
      "step": 2300
    },
    {
      "epoch": 0.5477593286913052,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2957,
      "step": 2301
    },
    {
      "epoch": 0.5479973814199846,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 2302
    },
    {
      "epoch": 0.5482354341486639,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 2303
    },
    {
      "epoch": 0.5484734868773433,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 2304
    },
    {
      "epoch": 0.5487115396060227,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3361,
      "step": 2305
    },
    {
      "epoch": 0.5489495923347021,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4429,
      "step": 2306
    },
    {
      "epoch": 0.5491876450633816,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.5671,
      "step": 2307
    },
    {
      "epoch": 0.549425697792061,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 2308
    },
    {
      "epoch": 0.5496637505207403,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 2309
    },
    {
      "epoch": 0.5499018032494197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3029,
      "step": 2310
    },
    {
      "epoch": 0.5501398559780991,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 2311
    },
    {
      "epoch": 0.5503779087067786,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3796,
      "step": 2312
    },
    {
      "epoch": 0.550615961435458,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 2313
    },
    {
      "epoch": 0.5508540141641374,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3845,
      "step": 2314
    },
    {
      "epoch": 0.5510920668928168,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 2315
    },
    {
      "epoch": 0.5513301196214961,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 2316
    },
    {
      "epoch": 0.5515681723501755,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 2317
    },
    {
      "epoch": 0.551806225078855,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 2318
    },
    {
      "epoch": 0.5520442778075344,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 2319
    },
    {
      "epoch": 0.5522823305362138,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 2320
    },
    {
      "epoch": 0.5525203832648932,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 2321
    },
    {
      "epoch": 0.5527584359935725,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 2322
    },
    {
      "epoch": 0.5529964887222519,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 2323
    },
    {
      "epoch": 0.5532345414509314,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 2324
    },
    {
      "epoch": 0.5534725941796108,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 2325
    },
    {
      "epoch": 0.5537106469082902,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2772,
      "step": 2326
    },
    {
      "epoch": 0.5539486996369696,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 2327
    },
    {
      "epoch": 0.554186752365649,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 2328
    },
    {
      "epoch": 0.5544248050943283,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 2329
    },
    {
      "epoch": 0.5546628578230078,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 2330
    },
    {
      "epoch": 0.5549009105516872,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 2331
    },
    {
      "epoch": 0.5551389632803666,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 2332
    },
    {
      "epoch": 0.555377016009046,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4834,
      "step": 2333
    },
    {
      "epoch": 0.5556150687377254,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 2334
    },
    {
      "epoch": 0.5558531214664048,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 2335
    },
    {
      "epoch": 0.5560911741950842,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3274,
      "step": 2336
    },
    {
      "epoch": 0.5563292269237636,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 2337
    },
    {
      "epoch": 0.556567279652443,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 2338
    },
    {
      "epoch": 0.5568053323811224,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4711,
      "step": 2339
    },
    {
      "epoch": 0.5570433851098018,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 2340
    },
    {
      "epoch": 0.5572814378384813,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 2341
    },
    {
      "epoch": 0.5575194905671607,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 2342
    },
    {
      "epoch": 0.55775754329584,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 2343
    },
    {
      "epoch": 0.5579955960245194,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 2344
    },
    {
      "epoch": 0.5582336487531988,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3274,
      "step": 2345
    },
    {
      "epoch": 0.5584717014818782,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 2346
    },
    {
      "epoch": 0.5587097542105577,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 2347
    },
    {
      "epoch": 0.5589478069392371,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 2348
    },
    {
      "epoch": 0.5591858596679165,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 2349
    },
    {
      "epoch": 0.5594239123965958,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4484,
      "step": 2350
    },
    {
      "epoch": 0.5596619651252752,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 2351
    },
    {
      "epoch": 0.5599000178539546,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 2352
    },
    {
      "epoch": 0.5601380705826341,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.451,
      "step": 2353
    },
    {
      "epoch": 0.5603761233113135,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 2354
    },
    {
      "epoch": 0.5606141760399929,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 2355
    },
    {
      "epoch": 0.5608522287686722,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 2356
    },
    {
      "epoch": 0.5610902814973516,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.5826,
      "step": 2357
    },
    {
      "epoch": 0.561328334226031,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 2358
    },
    {
      "epoch": 0.5615663869547105,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 2359
    },
    {
      "epoch": 0.5618044396833899,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 2360
    },
    {
      "epoch": 0.5620424924120693,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 2361
    },
    {
      "epoch": 0.5622805451407487,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 2362
    },
    {
      "epoch": 0.562518597869428,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.5341,
      "step": 2363
    },
    {
      "epoch": 0.5627566505981075,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 2364
    },
    {
      "epoch": 0.5629947033267869,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 2365
    },
    {
      "epoch": 0.5632327560554663,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 2366
    },
    {
      "epoch": 0.5634708087841457,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4645,
      "step": 2367
    },
    {
      "epoch": 0.5637088615128251,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 2368
    },
    {
      "epoch": 0.5639469142415044,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 2369
    },
    {
      "epoch": 0.5641849669701839,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.6404,
      "step": 2370
    },
    {
      "epoch": 0.5644230196988633,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 2371
    },
    {
      "epoch": 0.5646610724275427,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 2372
    },
    {
      "epoch": 0.5648991251562221,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 2373
    },
    {
      "epoch": 0.5651371778849015,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 2374
    },
    {
      "epoch": 0.5653752306135809,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 2375
    },
    {
      "epoch": 0.5656132833422604,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5376,
      "step": 2376
    },
    {
      "epoch": 0.5658513360709397,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 2377
    },
    {
      "epoch": 0.5660893887996191,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.4712,
      "step": 2378
    },
    {
      "epoch": 0.5663274415282985,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 2379
    },
    {
      "epoch": 0.5665654942569779,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 2380
    },
    {
      "epoch": 0.5668035469856573,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4556,
      "step": 2381
    },
    {
      "epoch": 0.5670415997143368,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2809,
      "step": 2382
    },
    {
      "epoch": 0.5672796524430161,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.537,
      "step": 2383
    },
    {
      "epoch": 0.5675177051716955,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 2384
    },
    {
      "epoch": 0.5677557579003749,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 2385
    },
    {
      "epoch": 0.5679938106290543,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 2386
    },
    {
      "epoch": 0.5682318633577338,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 2387
    },
    {
      "epoch": 0.5684699160864132,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 2388
    },
    {
      "epoch": 0.5687079688150926,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 2389
    },
    {
      "epoch": 0.5689460215437719,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 2390
    },
    {
      "epoch": 0.5691840742724513,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 2391
    },
    {
      "epoch": 0.5694221270011307,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 2392
    },
    {
      "epoch": 0.5696601797298102,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 2393
    },
    {
      "epoch": 0.5698982324584896,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2647,
      "step": 2394
    },
    {
      "epoch": 0.570136285187169,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 2395
    },
    {
      "epoch": 0.5703743379158484,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 2396
    },
    {
      "epoch": 0.5706123906445277,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 2397
    },
    {
      "epoch": 0.5708504433732071,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 2398
    },
    {
      "epoch": 0.5710884961018866,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 2399
    },
    {
      "epoch": 0.571326548830566,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3166,
      "step": 2400
    },
    {
      "epoch": 0.571326548830566,
      "eval_runtime": 0.5173,
      "eval_samples_per_second": 1.933,
      "eval_steps_per_second": 1.933,
      "step": 2400
    },
    {
      "epoch": 0.5715646015592454,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 2401
    },
    {
      "epoch": 0.5718026542879248,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3075,
      "step": 2402
    },
    {
      "epoch": 0.5720407070166041,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3023,
      "step": 2403
    },
    {
      "epoch": 0.5722787597452835,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 2404
    },
    {
      "epoch": 0.572516812473963,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 2405
    },
    {
      "epoch": 0.5727548652026424,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 2406
    },
    {
      "epoch": 0.5729929179313218,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3407,
      "step": 2407
    },
    {
      "epoch": 0.5732309706600012,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.4096,
      "step": 2408
    },
    {
      "epoch": 0.5734690233886806,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.5959,
      "step": 2409
    },
    {
      "epoch": 0.57370707611736,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 2410
    },
    {
      "epoch": 0.5739451288460394,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 2411
    },
    {
      "epoch": 0.5741831815747188,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4745,
      "step": 2412
    },
    {
      "epoch": 0.5744212343033982,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 2413
    },
    {
      "epoch": 0.5746592870320776,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 2414
    },
    {
      "epoch": 0.574897339760757,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 2415
    },
    {
      "epoch": 0.5751353924894365,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.426,
      "step": 2416
    },
    {
      "epoch": 0.5753734452181158,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5174,
      "step": 2417
    },
    {
      "epoch": 0.5756114979467952,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 2418
    },
    {
      "epoch": 0.5758495506754746,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5531,
      "step": 2419
    },
    {
      "epoch": 0.576087603404154,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2391,
      "step": 2420
    },
    {
      "epoch": 0.5763256561328334,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 2421
    },
    {
      "epoch": 0.5765637088615129,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 2422
    },
    {
      "epoch": 0.5768017615901923,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 2423
    },
    {
      "epoch": 0.5770398143188716,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 2424
    },
    {
      "epoch": 0.577277867047551,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 2425
    },
    {
      "epoch": 0.5775159197762304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3879,
      "step": 2426
    },
    {
      "epoch": 0.5777539725049098,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 2427
    },
    {
      "epoch": 0.5779920252335893,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 2428
    },
    {
      "epoch": 0.5782300779622687,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 2429
    },
    {
      "epoch": 0.578468130690948,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.404,
      "step": 2430
    },
    {
      "epoch": 0.5787061834196274,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 2431
    },
    {
      "epoch": 0.5789442361483068,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 2432
    },
    {
      "epoch": 0.5791822888769863,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4887,
      "step": 2433
    },
    {
      "epoch": 0.5794203416056657,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3591,
      "step": 2434
    },
    {
      "epoch": 0.5796583943343451,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 2435
    },
    {
      "epoch": 0.5798964470630245,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 2436
    },
    {
      "epoch": 0.5801344997917038,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3052,
      "step": 2437
    },
    {
      "epoch": 0.5803725525203832,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 2438
    },
    {
      "epoch": 0.5806106052490627,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4281,
      "step": 2439
    },
    {
      "epoch": 0.5808486579777421,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 2440
    },
    {
      "epoch": 0.5810867107064215,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 2441
    },
    {
      "epoch": 0.5813247634351009,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3094,
      "step": 2442
    },
    {
      "epoch": 0.5815628161637802,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 2443
    },
    {
      "epoch": 0.5818008688924596,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 2444
    },
    {
      "epoch": 0.5820389216211391,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4852,
      "step": 2445
    },
    {
      "epoch": 0.5822769743498185,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 2446
    },
    {
      "epoch": 0.5825150270784979,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 2447
    },
    {
      "epoch": 0.5827530798071773,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 2448
    },
    {
      "epoch": 0.5829911325358567,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3995,
      "step": 2449
    },
    {
      "epoch": 0.583229185264536,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4467,
      "step": 2450
    },
    {
      "epoch": 0.5834672379932155,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4654,
      "step": 2451
    },
    {
      "epoch": 0.5837052907218949,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 2452
    },
    {
      "epoch": 0.5839433434505743,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 2453
    },
    {
      "epoch": 0.5841813961792537,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2973,
      "step": 2454
    },
    {
      "epoch": 0.5844194489079331,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 2455
    },
    {
      "epoch": 0.5846575016366126,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 2456
    },
    {
      "epoch": 0.584895554365292,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 2457
    },
    {
      "epoch": 0.5851336070939713,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 2458
    },
    {
      "epoch": 0.5853716598226507,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 2459
    },
    {
      "epoch": 0.5856097125513301,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 2460
    },
    {
      "epoch": 0.5858477652800095,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 2461
    },
    {
      "epoch": 0.586085818008689,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 2462
    },
    {
      "epoch": 0.5863238707373684,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 2463
    },
    {
      "epoch": 0.5865619234660477,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 2464
    },
    {
      "epoch": 0.5867999761947271,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 2465
    },
    {
      "epoch": 0.5870380289234065,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 2466
    },
    {
      "epoch": 0.5872760816520859,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 2467
    },
    {
      "epoch": 0.5875141343807654,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 2468
    },
    {
      "epoch": 0.5877521871094448,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 2469
    },
    {
      "epoch": 0.5879902398381242,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2649,
      "step": 2470
    },
    {
      "epoch": 0.5882282925668035,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 2471
    },
    {
      "epoch": 0.5884663452954829,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 2472
    },
    {
      "epoch": 0.5887043980241623,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 2473
    },
    {
      "epoch": 0.5889424507528418,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 2474
    },
    {
      "epoch": 0.5891805034815212,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 2475
    },
    {
      "epoch": 0.5894185562102006,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2835,
      "step": 2476
    },
    {
      "epoch": 0.58965660893888,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 2477
    },
    {
      "epoch": 0.5898946616675593,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.554,
      "step": 2478
    },
    {
      "epoch": 0.5901327143962388,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 2479
    },
    {
      "epoch": 0.5903707671249182,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 2480
    },
    {
      "epoch": 0.5906088198535976,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 2481
    },
    {
      "epoch": 0.590846872582277,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 2482
    },
    {
      "epoch": 0.5910849253109564,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4641,
      "step": 2483
    },
    {
      "epoch": 0.5913229780396357,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 2484
    },
    {
      "epoch": 0.5915610307683152,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3721,
      "step": 2485
    },
    {
      "epoch": 0.5917990834969946,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 2486
    },
    {
      "epoch": 0.592037136225674,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 2487
    },
    {
      "epoch": 0.5922751889543534,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 2488
    },
    {
      "epoch": 0.5925132416830328,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 2489
    },
    {
      "epoch": 0.5927512944117121,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 2490
    },
    {
      "epoch": 0.5929893471403916,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3885,
      "step": 2491
    },
    {
      "epoch": 0.593227399869071,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5117,
      "step": 2492
    },
    {
      "epoch": 0.5934654525977504,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2895,
      "step": 2493
    },
    {
      "epoch": 0.5937035053264298,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3962,
      "step": 2494
    },
    {
      "epoch": 0.5939415580551092,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 2495
    },
    {
      "epoch": 0.5941796107837886,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 2496
    },
    {
      "epoch": 0.594417663512468,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3323,
      "step": 2497
    },
    {
      "epoch": 0.5946557162411474,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 2498
    },
    {
      "epoch": 0.5948937689698268,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 2499
    },
    {
      "epoch": 0.5951318216985062,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 2500
    },
    {
      "epoch": 0.5951318216985062,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 2500
    },
    {
      "epoch": 0.5953698744271856,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 2501
    },
    {
      "epoch": 0.595607927155865,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 2502
    },
    {
      "epoch": 0.5958459798845445,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 2503
    },
    {
      "epoch": 0.5960840326132238,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4137,
      "step": 2504
    },
    {
      "epoch": 0.5963220853419032,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4238,
      "step": 2505
    },
    {
      "epoch": 0.5965601380705826,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 2506
    },
    {
      "epoch": 0.596798190799262,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 2507
    },
    {
      "epoch": 0.5970362435279415,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 2508
    },
    {
      "epoch": 0.5972742962566209,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 2509
    },
    {
      "epoch": 0.5975123489853003,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.4147,
      "step": 2510
    },
    {
      "epoch": 0.5977504017139796,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 2511
    },
    {
      "epoch": 0.597988454442659,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 2512
    },
    {
      "epoch": 0.5982265071713384,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 2513
    },
    {
      "epoch": 0.5984645599000179,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 2514
    },
    {
      "epoch": 0.5987026126286973,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.401,
      "step": 2515
    },
    {
      "epoch": 0.5989406653573767,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.6005,
      "step": 2516
    },
    {
      "epoch": 0.599178718086056,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 2517
    },
    {
      "epoch": 0.5994167708147354,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3912,
      "step": 2518
    },
    {
      "epoch": 0.5996548235434148,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.394,
      "step": 2519
    },
    {
      "epoch": 0.5998928762720943,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 2520
    },
    {
      "epoch": 0.6001309290007737,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4162,
      "step": 2521
    },
    {
      "epoch": 0.6003689817294531,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3674,
      "step": 2522
    },
    {
      "epoch": 0.6006070344581325,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5469,
      "step": 2523
    },
    {
      "epoch": 0.6008450871868118,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 2524
    },
    {
      "epoch": 0.6010831399154912,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 2525
    },
    {
      "epoch": 0.6013211926441707,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 2526
    },
    {
      "epoch": 0.6015592453728501,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 2527
    },
    {
      "epoch": 0.6017972981015295,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 2528
    },
    {
      "epoch": 0.6020353508302089,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 2529
    },
    {
      "epoch": 0.6022734035588883,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 2530
    },
    {
      "epoch": 0.6025114562875677,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 2531
    },
    {
      "epoch": 0.6027495090162471,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 2532
    },
    {
      "epoch": 0.6029875617449265,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 2533
    },
    {
      "epoch": 0.6032256144736059,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 2534
    },
    {
      "epoch": 0.6034636672022853,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3473,
      "step": 2535
    },
    {
      "epoch": 0.6037017199309647,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 2536
    },
    {
      "epoch": 0.6039397726596442,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3431,
      "step": 2537
    },
    {
      "epoch": 0.6041778253883235,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 2538
    },
    {
      "epoch": 0.6044158781170029,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 2539
    },
    {
      "epoch": 0.6046539308456823,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 2540
    },
    {
      "epoch": 0.6048919835743617,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.3226,
      "step": 2541
    },
    {
      "epoch": 0.6051300363030411,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 2542
    },
    {
      "epoch": 0.6053680890317206,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 2543
    },
    {
      "epoch": 0.6056061417604,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 2544
    },
    {
      "epoch": 0.6058441944890793,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 2545
    },
    {
      "epoch": 0.6060822472177587,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 2546
    },
    {
      "epoch": 0.6063202999464381,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 2547
    },
    {
      "epoch": 0.6065583526751175,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.5055,
      "step": 2548
    },
    {
      "epoch": 0.606796405403797,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3443,
      "step": 2549
    },
    {
      "epoch": 0.6070344581324764,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3764,
      "step": 2550
    },
    {
      "epoch": 0.6072725108611557,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 2551
    },
    {
      "epoch": 0.6075105635898351,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 2552
    },
    {
      "epoch": 0.6077486163185145,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 2553
    },
    {
      "epoch": 0.607986669047194,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 2554
    },
    {
      "epoch": 0.6082247217758734,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 2555
    },
    {
      "epoch": 0.6084627745045528,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 2556
    },
    {
      "epoch": 0.6087008272332322,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 2557
    },
    {
      "epoch": 0.6089388799619115,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 2558
    },
    {
      "epoch": 0.6091769326905909,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 2559
    },
    {
      "epoch": 0.6094149854192704,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.483,
      "step": 2560
    },
    {
      "epoch": 0.6096530381479498,
      "grad_norm": 2.671875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 2561
    },
    {
      "epoch": 0.6098910908766292,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 2562
    },
    {
      "epoch": 0.6101291436053086,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 2563
    },
    {
      "epoch": 0.610367196333988,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 2564
    },
    {
      "epoch": 0.6106052490626673,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 2565
    },
    {
      "epoch": 0.6108433017913468,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 2566
    },
    {
      "epoch": 0.6110813545200262,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 2567
    },
    {
      "epoch": 0.6113194072487056,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 2568
    },
    {
      "epoch": 0.611557459977385,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 2569
    },
    {
      "epoch": 0.6117955127060644,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 2570
    },
    {
      "epoch": 0.6120335654347437,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 2571
    },
    {
      "epoch": 0.6122716181634232,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 2572
    },
    {
      "epoch": 0.6125096708921026,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 2573
    },
    {
      "epoch": 0.612747723620782,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 2574
    },
    {
      "epoch": 0.6129857763494614,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 2575
    },
    {
      "epoch": 0.6132238290781408,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2855,
      "step": 2576
    },
    {
      "epoch": 0.6134618818068203,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.355,
      "step": 2577
    },
    {
      "epoch": 0.6136999345354996,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 2578
    },
    {
      "epoch": 0.613937987264179,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3751,
      "step": 2579
    },
    {
      "epoch": 0.6141760399928584,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 2580
    },
    {
      "epoch": 0.6144140927215378,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 2581
    },
    {
      "epoch": 0.6146521454502172,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3708,
      "step": 2582
    },
    {
      "epoch": 0.6148901981788967,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8236,
      "step": 2583
    },
    {
      "epoch": 0.6151282509075761,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3318,
      "step": 2584
    },
    {
      "epoch": 0.6153663036362554,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 2585
    },
    {
      "epoch": 0.6156043563649348,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 2586
    },
    {
      "epoch": 0.6158424090936142,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2614,
      "step": 2587
    },
    {
      "epoch": 0.6160804618222936,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 2588
    },
    {
      "epoch": 0.6163185145509731,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3001,
      "step": 2589
    },
    {
      "epoch": 0.6165565672796525,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5246,
      "step": 2590
    },
    {
      "epoch": 0.6167946200083319,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 2591
    },
    {
      "epoch": 0.6170326727370112,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 2592
    },
    {
      "epoch": 0.6172707254656906,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 2593
    },
    {
      "epoch": 0.61750877819437,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 2594
    },
    {
      "epoch": 0.6177468309230495,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 2595
    },
    {
      "epoch": 0.6179848836517289,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 2596
    },
    {
      "epoch": 0.6182229363804083,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2816,
      "step": 2597
    },
    {
      "epoch": 0.6184609891090876,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 2598
    },
    {
      "epoch": 0.618699041837767,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 2599
    },
    {
      "epoch": 0.6189370945664465,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 2600
    },
    {
      "epoch": 0.6189370945664465,
      "eval_runtime": 0.51,
      "eval_samples_per_second": 1.961,
      "eval_steps_per_second": 1.961,
      "step": 2600
    },
    {
      "epoch": 0.6191751472951259,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 2601
    },
    {
      "epoch": 0.6194132000238053,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 2602
    },
    {
      "epoch": 0.6196512527524847,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3571,
      "step": 2603
    },
    {
      "epoch": 0.619889305481164,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 2604
    },
    {
      "epoch": 0.6201273582098434,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 2605
    },
    {
      "epoch": 0.6203654109385229,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.6442,
      "step": 2606
    },
    {
      "epoch": 0.6206034636672023,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 2607
    },
    {
      "epoch": 0.6208415163958817,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 2608
    },
    {
      "epoch": 0.6210795691245611,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 2609
    },
    {
      "epoch": 0.6213176218532405,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 2610
    },
    {
      "epoch": 0.6215556745819198,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 2611
    },
    {
      "epoch": 0.6217937273105993,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4408,
      "step": 2612
    },
    {
      "epoch": 0.6220317800392787,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.401,
      "step": 2613
    },
    {
      "epoch": 0.6222698327679581,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3785,
      "step": 2614
    },
    {
      "epoch": 0.6225078854966375,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.5327,
      "step": 2615
    },
    {
      "epoch": 0.6227459382253169,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 2616
    },
    {
      "epoch": 0.6229839909539963,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 2617
    },
    {
      "epoch": 0.6232220436826758,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 2618
    },
    {
      "epoch": 0.6234600964113551,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 2619
    },
    {
      "epoch": 0.6236981491400345,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 2620
    },
    {
      "epoch": 0.6239362018687139,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4455,
      "step": 2621
    },
    {
      "epoch": 0.6241742545973933,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 2622
    },
    {
      "epoch": 0.6244123073260728,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4936,
      "step": 2623
    },
    {
      "epoch": 0.6246503600547522,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 2624
    },
    {
      "epoch": 0.6248884127834315,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 2625
    },
    {
      "epoch": 0.6251264655121109,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 2626
    },
    {
      "epoch": 0.6253645182407903,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 2627
    },
    {
      "epoch": 0.6256025709694697,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4343,
      "step": 2628
    },
    {
      "epoch": 0.6258406236981492,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 2629
    },
    {
      "epoch": 0.6260786764268286,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 2630
    },
    {
      "epoch": 0.626316729155508,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.4375,
      "step": 2631
    },
    {
      "epoch": 0.6265547818841873,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 2632
    },
    {
      "epoch": 0.6267928346128667,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 2633
    },
    {
      "epoch": 0.6270308873415461,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 2634
    },
    {
      "epoch": 0.6272689400702256,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 2635
    },
    {
      "epoch": 0.627506992798905,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 2636
    },
    {
      "epoch": 0.6277450455275844,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 2637
    },
    {
      "epoch": 0.6279830982562637,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2621,
      "step": 2638
    },
    {
      "epoch": 0.6282211509849431,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 2639
    },
    {
      "epoch": 0.6284592037136225,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.6351,
      "step": 2640
    },
    {
      "epoch": 0.628697256442302,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.447,
      "step": 2641
    },
    {
      "epoch": 0.6289353091709814,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 2642
    },
    {
      "epoch": 0.6291733618996608,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 2643
    },
    {
      "epoch": 0.6294114146283402,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 2644
    },
    {
      "epoch": 0.6296494673570195,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 2645
    },
    {
      "epoch": 0.629887520085699,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 2646
    },
    {
      "epoch": 0.6301255728143784,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2689,
      "step": 2647
    },
    {
      "epoch": 0.6303636255430578,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 2648
    },
    {
      "epoch": 0.6306016782717372,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3802,
      "step": 2649
    },
    {
      "epoch": 0.6308397310004166,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3517,
      "step": 2650
    },
    {
      "epoch": 0.631077783729096,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.5992,
      "step": 2651
    },
    {
      "epoch": 0.6313158364577754,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.437,
      "step": 2652
    },
    {
      "epoch": 0.6315538891864548,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 2653
    },
    {
      "epoch": 0.6317919419151342,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 2654
    },
    {
      "epoch": 0.6320299946438136,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4049,
      "step": 2655
    },
    {
      "epoch": 0.632268047372493,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 2656
    },
    {
      "epoch": 0.6325061001011724,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4043,
      "step": 2657
    },
    {
      "epoch": 0.6327441528298519,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 2658
    },
    {
      "epoch": 0.6329822055585312,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.428,
      "step": 2659
    },
    {
      "epoch": 0.6332202582872106,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4005,
      "step": 2660
    },
    {
      "epoch": 0.63345831101589,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3792,
      "step": 2661
    },
    {
      "epoch": 0.6336963637445694,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 2662
    },
    {
      "epoch": 0.6339344164732488,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.6701,
      "step": 2663
    },
    {
      "epoch": 0.6341724692019283,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 2664
    },
    {
      "epoch": 0.6344105219306077,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 2665
    },
    {
      "epoch": 0.634648574659287,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4384,
      "step": 2666
    },
    {
      "epoch": 0.6348866273879664,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 2667
    },
    {
      "epoch": 0.6351246801166458,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 2668
    },
    {
      "epoch": 0.6353627328453252,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 2669
    },
    {
      "epoch": 0.6356007855740047,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 2670
    },
    {
      "epoch": 0.6358388383026841,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 2671
    },
    {
      "epoch": 0.6360768910313634,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 2672
    },
    {
      "epoch": 0.6363149437600428,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 2673
    },
    {
      "epoch": 0.6365529964887222,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 2674
    },
    {
      "epoch": 0.6367910492174017,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 2675
    },
    {
      "epoch": 0.6370291019460811,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 2676
    },
    {
      "epoch": 0.6372671546747605,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 2677
    },
    {
      "epoch": 0.6375052074034399,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3744,
      "step": 2678
    },
    {
      "epoch": 0.6377432601321192,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 2679
    },
    {
      "epoch": 0.6379813128607986,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 2680
    },
    {
      "epoch": 0.6382193655894781,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2555,
      "step": 2681
    },
    {
      "epoch": 0.6384574183181575,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 2682
    },
    {
      "epoch": 0.6386954710468369,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 2683
    },
    {
      "epoch": 0.6389335237755163,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 2684
    },
    {
      "epoch": 0.6391715765041956,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 2685
    },
    {
      "epoch": 0.639409629232875,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 2686
    },
    {
      "epoch": 0.6396476819615545,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 2687
    },
    {
      "epoch": 0.6398857346902339,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4637,
      "step": 2688
    },
    {
      "epoch": 0.6401237874189133,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 2689
    },
    {
      "epoch": 0.6403618401475927,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 2690
    },
    {
      "epoch": 0.6405998928762721,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 2691
    },
    {
      "epoch": 0.6408379456049514,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 2692
    },
    {
      "epoch": 0.6410759983336309,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 2693
    },
    {
      "epoch": 0.6413140510623103,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 2694
    },
    {
      "epoch": 0.6415521037909897,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 2695
    },
    {
      "epoch": 0.6417901565196691,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3178,
      "step": 2696
    },
    {
      "epoch": 0.6420282092483485,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 2697
    },
    {
      "epoch": 0.642266261977028,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 2698
    },
    {
      "epoch": 0.6425043147057073,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 2699
    },
    {
      "epoch": 0.6427423674343867,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 2700
    },
    {
      "epoch": 0.6427423674343867,
      "eval_runtime": 0.5064,
      "eval_samples_per_second": 1.975,
      "eval_steps_per_second": 1.975,
      "step": 2700
    },
    {
      "epoch": 0.6429804201630661,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.494,
      "step": 2701
    },
    {
      "epoch": 0.6432184728917455,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 2702
    },
    {
      "epoch": 0.6434565256204249,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3703,
      "step": 2703
    },
    {
      "epoch": 0.6436945783491044,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3739,
      "step": 2704
    },
    {
      "epoch": 0.6439326310777838,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 2705
    },
    {
      "epoch": 0.6441706838064631,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3486,
      "step": 2706
    },
    {
      "epoch": 0.6444087365351425,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 2707
    },
    {
      "epoch": 0.6446467892638219,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 2708
    },
    {
      "epoch": 0.6448848419925013,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 2709
    },
    {
      "epoch": 0.6451228947211808,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2948,
      "step": 2710
    },
    {
      "epoch": 0.6453609474498602,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.7267,
      "step": 2711
    },
    {
      "epoch": 0.6455990001785396,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 2712
    },
    {
      "epoch": 0.6458370529072189,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 2713
    },
    {
      "epoch": 0.6460751056358983,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 2714
    },
    {
      "epoch": 0.6463131583645777,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 2715
    },
    {
      "epoch": 0.6465512110932572,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 2716
    },
    {
      "epoch": 0.6467892638219366,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3742,
      "step": 2717
    },
    {
      "epoch": 0.647027316550616,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.374,
      "step": 2718
    },
    {
      "epoch": 0.6472653692792953,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 2719
    },
    {
      "epoch": 0.6475034220079747,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 2720
    },
    {
      "epoch": 0.6477414747366542,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.342,
      "step": 2721
    },
    {
      "epoch": 0.6479795274653336,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 2722
    },
    {
      "epoch": 0.648217580194013,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 2723
    },
    {
      "epoch": 0.6484556329226924,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 2724
    },
    {
      "epoch": 0.6486936856513718,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 2725
    },
    {
      "epoch": 0.6489317383800511,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 2726
    },
    {
      "epoch": 0.6491697911087306,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 2727
    },
    {
      "epoch": 0.64940784383741,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4815,
      "step": 2728
    },
    {
      "epoch": 0.6496458965660894,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 2729
    },
    {
      "epoch": 0.6498839492947688,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 2730
    },
    {
      "epoch": 0.6501220020234482,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 2731
    },
    {
      "epoch": 0.6503600547521275,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 2732
    },
    {
      "epoch": 0.650598107480807,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 2733
    },
    {
      "epoch": 0.6508361602094864,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3587,
      "step": 2734
    },
    {
      "epoch": 0.6510742129381658,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 2735
    },
    {
      "epoch": 0.6513122656668452,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 2736
    },
    {
      "epoch": 0.6515503183955246,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 2737
    },
    {
      "epoch": 0.651788371124204,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 2738
    },
    {
      "epoch": 0.6520264238528835,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 2739
    },
    {
      "epoch": 0.6522644765815628,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 2740
    },
    {
      "epoch": 0.6525025293102422,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 2741
    },
    {
      "epoch": 0.6527405820389216,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3945,
      "step": 2742
    },
    {
      "epoch": 0.652978634767601,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 2743
    },
    {
      "epoch": 0.6532166874962805,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 2744
    },
    {
      "epoch": 0.6534547402249599,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.5306,
      "step": 2745
    },
    {
      "epoch": 0.6536927929536392,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 2746
    },
    {
      "epoch": 0.6539308456823186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 2747
    },
    {
      "epoch": 0.654168898410998,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 2748
    },
    {
      "epoch": 0.6544069511396774,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 2749
    },
    {
      "epoch": 0.6546450038683569,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 2750
    },
    {
      "epoch": 0.6548830565970363,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 2751
    },
    {
      "epoch": 0.6551211093257157,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 2752
    },
    {
      "epoch": 0.655359162054395,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 2753
    },
    {
      "epoch": 0.6555972147830744,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 2754
    },
    {
      "epoch": 0.6558352675117538,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 2755
    },
    {
      "epoch": 0.6560733202404333,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 2756
    },
    {
      "epoch": 0.6563113729691127,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.687,
      "step": 2757
    },
    {
      "epoch": 0.6565494256977921,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.6769,
      "step": 2758
    },
    {
      "epoch": 0.6567874784264714,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 2759
    },
    {
      "epoch": 0.6570255311551508,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 2760
    },
    {
      "epoch": 0.6572635838838302,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 2761
    },
    {
      "epoch": 0.6575016366125097,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 2762
    },
    {
      "epoch": 0.6577396893411891,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 2763
    },
    {
      "epoch": 0.6579777420698685,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 2764
    },
    {
      "epoch": 0.6582157947985479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3769,
      "step": 2765
    },
    {
      "epoch": 0.6584538475272272,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 2766
    },
    {
      "epoch": 0.6586919002559067,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2466,
      "step": 2767
    },
    {
      "epoch": 0.6589299529845861,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 2768
    },
    {
      "epoch": 0.6591680057132655,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 2769
    },
    {
      "epoch": 0.6594060584419449,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 2770
    },
    {
      "epoch": 0.6596441111706243,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3905,
      "step": 2771
    },
    {
      "epoch": 0.6598821638993037,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3863,
      "step": 2772
    },
    {
      "epoch": 0.6601202166279831,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 2773
    },
    {
      "epoch": 0.6603582693566625,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 2774
    },
    {
      "epoch": 0.6605963220853419,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2604,
      "step": 2775
    },
    {
      "epoch": 0.6608343748140213,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 2776
    },
    {
      "epoch": 0.6610724275427007,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 2777
    },
    {
      "epoch": 0.6613104802713801,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.3538,
      "step": 2778
    },
    {
      "epoch": 0.6615485330000596,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 2779
    },
    {
      "epoch": 0.6617865857287389,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 2780
    },
    {
      "epoch": 0.6620246384574183,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4847,
      "step": 2781
    },
    {
      "epoch": 0.6622626911860977,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 2782
    },
    {
      "epoch": 0.6625007439147771,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 2783
    },
    {
      "epoch": 0.6627387966434565,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 2784
    },
    {
      "epoch": 0.662976849372136,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 2785
    },
    {
      "epoch": 0.6632149021008154,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 2786
    },
    {
      "epoch": 0.6634529548294947,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 2787
    },
    {
      "epoch": 0.6636910075581741,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 2788
    },
    {
      "epoch": 0.6639290602868535,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 2789
    },
    {
      "epoch": 0.664167113015533,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 2790
    },
    {
      "epoch": 0.6644051657442124,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2269,
      "step": 2791
    },
    {
      "epoch": 0.6646432184728918,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4788,
      "step": 2792
    },
    {
      "epoch": 0.6648812712015711,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 2793
    },
    {
      "epoch": 0.6651193239302505,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 2794
    },
    {
      "epoch": 0.6653573766589299,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 2795
    },
    {
      "epoch": 0.6655954293876094,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 2796
    },
    {
      "epoch": 0.6658334821162888,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 2797
    },
    {
      "epoch": 0.6660715348449682,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 2798
    },
    {
      "epoch": 0.6663095875736476,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2558,
      "step": 2799
    },
    {
      "epoch": 0.6665476403023269,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 2800
    },
    {
      "epoch": 0.6665476403023269,
      "eval_runtime": 0.5251,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 2800
    },
    {
      "epoch": 0.6667856930310063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 2801
    },
    {
      "epoch": 0.6670237457596858,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 2802
    },
    {
      "epoch": 0.6672617984883652,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 2803
    },
    {
      "epoch": 0.6674998512170446,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 2804
    },
    {
      "epoch": 0.667737903945724,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.5395,
      "step": 2805
    },
    {
      "epoch": 0.6679759566744033,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 2806
    },
    {
      "epoch": 0.6682140094030827,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 2807
    },
    {
      "epoch": 0.6684520621317622,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4558,
      "step": 2808
    },
    {
      "epoch": 0.6686901148604416,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 2809
    },
    {
      "epoch": 0.668928167589121,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 2810
    },
    {
      "epoch": 0.6691662203178004,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 2811
    },
    {
      "epoch": 0.6694042730464798,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 2812
    },
    {
      "epoch": 0.6696423257751593,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 2813
    },
    {
      "epoch": 0.6698803785038386,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3676,
      "step": 2814
    },
    {
      "epoch": 0.670118431232518,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3045,
      "step": 2815
    },
    {
      "epoch": 0.6703564839611974,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 2816
    },
    {
      "epoch": 0.6705945366898768,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 2817
    },
    {
      "epoch": 0.6708325894185562,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 2818
    },
    {
      "epoch": 0.6710706421472357,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.6284,
      "step": 2819
    },
    {
      "epoch": 0.671308694875915,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 2820
    },
    {
      "epoch": 0.6715467476045944,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 2821
    },
    {
      "epoch": 0.6717848003332738,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 2822
    },
    {
      "epoch": 0.6720228530619532,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 2823
    },
    {
      "epoch": 0.6722609057906326,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 2824
    },
    {
      "epoch": 0.6724989585193121,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 2825
    },
    {
      "epoch": 0.6727370112479915,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 2826
    },
    {
      "epoch": 0.6729750639766708,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.4121,
      "step": 2827
    },
    {
      "epoch": 0.6732131167053502,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 2828
    },
    {
      "epoch": 0.6734511694340296,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2694,
      "step": 2829
    },
    {
      "epoch": 0.673689222162709,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 2830
    },
    {
      "epoch": 0.6739272748913885,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 2831
    },
    {
      "epoch": 0.6741653276200679,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 2832
    },
    {
      "epoch": 0.6744033803487473,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 2833
    },
    {
      "epoch": 0.6746414330774266,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 2834
    },
    {
      "epoch": 0.674879485806106,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 2835
    },
    {
      "epoch": 0.6751175385347854,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 2836
    },
    {
      "epoch": 0.6753555912634649,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 2837
    },
    {
      "epoch": 0.6755936439921443,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 2838
    },
    {
      "epoch": 0.6758316967208237,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 2839
    },
    {
      "epoch": 0.676069749449503,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 2840
    },
    {
      "epoch": 0.6763078021781824,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 2841
    },
    {
      "epoch": 0.6765458549068619,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 2842
    },
    {
      "epoch": 0.6767839076355413,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 2843
    },
    {
      "epoch": 0.6770219603642207,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 2844
    },
    {
      "epoch": 0.6772600130929001,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 2845
    },
    {
      "epoch": 0.6774980658215795,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 2846
    },
    {
      "epoch": 0.6777361185502588,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 2847
    },
    {
      "epoch": 0.6779741712789383,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 2848
    },
    {
      "epoch": 0.6782122240076177,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 2849
    },
    {
      "epoch": 0.6784502767362971,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 2850
    },
    {
      "epoch": 0.6786883294649765,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.5165,
      "step": 2851
    },
    {
      "epoch": 0.6789263821936559,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 2852
    },
    {
      "epoch": 0.6791644349223352,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3773,
      "step": 2853
    },
    {
      "epoch": 0.6794024876510147,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 2854
    },
    {
      "epoch": 0.6796405403796941,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 2855
    },
    {
      "epoch": 0.6798785931083735,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.2501,
      "step": 2856
    },
    {
      "epoch": 0.6801166458370529,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4676,
      "step": 2857
    },
    {
      "epoch": 0.6803546985657323,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 2858
    },
    {
      "epoch": 0.6805927512944117,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 2859
    },
    {
      "epoch": 0.6808308040230912,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 2860
    },
    {
      "epoch": 0.6810688567517705,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4862,
      "step": 2861
    },
    {
      "epoch": 0.6813069094804499,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4726,
      "step": 2862
    },
    {
      "epoch": 0.6815449622091293,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 2863
    },
    {
      "epoch": 0.6817830149378087,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 2864
    },
    {
      "epoch": 0.6820210676664882,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 2865
    },
    {
      "epoch": 0.6822591203951676,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 2866
    },
    {
      "epoch": 0.682497173123847,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 2867
    },
    {
      "epoch": 0.6827352258525263,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 2868
    },
    {
      "epoch": 0.6829732785812057,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 2869
    },
    {
      "epoch": 0.6832113313098851,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3397,
      "step": 2870
    },
    {
      "epoch": 0.6834493840385646,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 2871
    },
    {
      "epoch": 0.683687436767244,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 2872
    },
    {
      "epoch": 0.6839254894959234,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 2873
    },
    {
      "epoch": 0.6841635422246027,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 2874
    },
    {
      "epoch": 0.6844015949532821,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 2875
    },
    {
      "epoch": 0.6846396476819615,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 2876
    },
    {
      "epoch": 0.684877700410641,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 2877
    },
    {
      "epoch": 0.6851157531393204,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 2878
    },
    {
      "epoch": 0.6853538058679998,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 2879
    },
    {
      "epoch": 0.6855918585966791,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 2880
    },
    {
      "epoch": 0.6858299113253585,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4161,
      "step": 2881
    },
    {
      "epoch": 0.6860679640540379,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 2882
    },
    {
      "epoch": 0.6863060167827174,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 2883
    },
    {
      "epoch": 0.6865440695113968,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 2884
    },
    {
      "epoch": 0.6867821222400762,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4406,
      "step": 2885
    },
    {
      "epoch": 0.6870201749687556,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 2886
    },
    {
      "epoch": 0.6872582276974349,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 2887
    },
    {
      "epoch": 0.6874962804261144,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2986,
      "step": 2888
    },
    {
      "epoch": 0.6877343331547938,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 2889
    },
    {
      "epoch": 0.6879723858834732,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.497,
      "step": 2890
    },
    {
      "epoch": 0.6882104386121526,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 2891
    },
    {
      "epoch": 0.688448491340832,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3422,
      "step": 2892
    },
    {
      "epoch": 0.6886865440695114,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3227,
      "step": 2893
    },
    {
      "epoch": 0.6889245967981908,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 2894
    },
    {
      "epoch": 0.6891626495268702,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 2895
    },
    {
      "epoch": 0.6894007022555496,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3323,
      "step": 2896
    },
    {
      "epoch": 0.689638754984229,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 2897
    },
    {
      "epoch": 0.6898768077129084,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 2898
    },
    {
      "epoch": 0.6901148604415878,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 2899
    },
    {
      "epoch": 0.6903529131702673,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 2900
    },
    {
      "epoch": 0.6903529131702673,
      "eval_runtime": 0.5307,
      "eval_samples_per_second": 1.884,
      "eval_steps_per_second": 1.884,
      "step": 2900
    },
    {
      "epoch": 0.6905909658989466,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 2901
    },
    {
      "epoch": 0.690829018627626,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 2902
    },
    {
      "epoch": 0.6910670713563054,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.532,
      "step": 2903
    },
    {
      "epoch": 0.6913051240849848,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 2904
    },
    {
      "epoch": 0.6915431768136642,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 2905
    },
    {
      "epoch": 0.6917812295423437,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0396,
      "step": 2906
    },
    {
      "epoch": 0.692019282271023,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 2907
    },
    {
      "epoch": 0.6922573349997024,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 2908
    },
    {
      "epoch": 0.6924953877283818,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 2909
    },
    {
      "epoch": 0.6927334404570612,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 2910
    },
    {
      "epoch": 0.6929714931857407,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 2911
    },
    {
      "epoch": 0.6932095459144201,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.614,
      "step": 2912
    },
    {
      "epoch": 0.6934475986430995,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 2913
    },
    {
      "epoch": 0.6936856513717788,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3568,
      "step": 2914
    },
    {
      "epoch": 0.6939237041004582,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 2915
    },
    {
      "epoch": 0.6941617568291376,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 2916
    },
    {
      "epoch": 0.6943998095578171,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 2917
    },
    {
      "epoch": 0.6946378622864965,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 2918
    },
    {
      "epoch": 0.6948759150151759,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 2919
    },
    {
      "epoch": 0.6951139677438553,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 2920
    },
    {
      "epoch": 0.6953520204725346,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 2921
    },
    {
      "epoch": 0.695590073201214,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3389,
      "step": 2922
    },
    {
      "epoch": 0.6958281259298935,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 2923
    },
    {
      "epoch": 0.6960661786585729,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 2924
    },
    {
      "epoch": 0.6963042313872523,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 2925
    },
    {
      "epoch": 0.6965422841159317,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 2926
    },
    {
      "epoch": 0.696780336844611,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 2927
    },
    {
      "epoch": 0.6970183895732904,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 2928
    },
    {
      "epoch": 0.6972564423019699,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4357,
      "step": 2929
    },
    {
      "epoch": 0.6974944950306493,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 2930
    },
    {
      "epoch": 0.6977325477593287,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.4695,
      "step": 2931
    },
    {
      "epoch": 0.6979706004880081,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.5153,
      "step": 2932
    },
    {
      "epoch": 0.6982086532166875,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 2933
    },
    {
      "epoch": 0.698446705945367,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4321,
      "step": 2934
    },
    {
      "epoch": 0.6986847586740463,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 2935
    },
    {
      "epoch": 0.6989228114027257,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 2936
    },
    {
      "epoch": 0.6991608641314051,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 2937
    },
    {
      "epoch": 0.6993989168600845,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 2938
    },
    {
      "epoch": 0.6996369695887639,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 2939
    },
    {
      "epoch": 0.6998750223174434,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 2940
    },
    {
      "epoch": 0.7001130750461227,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 2941
    },
    {
      "epoch": 0.7003511277748021,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 2942
    },
    {
      "epoch": 0.7005891805034815,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 2943
    },
    {
      "epoch": 0.7008272332321609,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 2944
    },
    {
      "epoch": 0.7010652859608403,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.217,
      "step": 2945
    },
    {
      "epoch": 0.7013033386895198,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3862,
      "step": 2946
    },
    {
      "epoch": 0.7015413914181992,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 2947
    },
    {
      "epoch": 0.7017794441468785,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 2948
    },
    {
      "epoch": 0.7020174968755579,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 2949
    },
    {
      "epoch": 0.7022555496042373,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 2950
    },
    {
      "epoch": 0.7024936023329167,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 2951
    },
    {
      "epoch": 0.7027316550615962,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 2952
    },
    {
      "epoch": 0.7029697077902756,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 2953
    },
    {
      "epoch": 0.703207760518955,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 2954
    },
    {
      "epoch": 0.7034458132476343,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 2955
    },
    {
      "epoch": 0.7036838659763137,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 2956
    },
    {
      "epoch": 0.7039219187049932,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 2957
    },
    {
      "epoch": 0.7041599714336726,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 2958
    },
    {
      "epoch": 0.704398024162352,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4699,
      "step": 2959
    },
    {
      "epoch": 0.7046360768910314,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.29,
      "step": 2960
    },
    {
      "epoch": 0.7048741296197107,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 2961
    },
    {
      "epoch": 0.7051121823483901,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 2962
    },
    {
      "epoch": 0.7053502350770696,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 2963
    },
    {
      "epoch": 0.705588287805749,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 2964
    },
    {
      "epoch": 0.7058263405344284,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 2965
    },
    {
      "epoch": 0.7060643932631078,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 2966
    },
    {
      "epoch": 0.7063024459917872,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 2967
    },
    {
      "epoch": 0.7065404987204665,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 2968
    },
    {
      "epoch": 0.706778551449146,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 2969
    },
    {
      "epoch": 0.7070166041778254,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.35,
      "step": 2970
    },
    {
      "epoch": 0.7072546569065048,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 2971
    },
    {
      "epoch": 0.7074927096351842,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 2972
    },
    {
      "epoch": 0.7077307623638636,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 2973
    },
    {
      "epoch": 0.707968815092543,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 2974
    },
    {
      "epoch": 0.7082068678212224,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 2975
    },
    {
      "epoch": 0.7084449205499018,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4148,
      "step": 2976
    },
    {
      "epoch": 0.7086829732785812,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 2977
    },
    {
      "epoch": 0.7089210260072606,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 2978
    },
    {
      "epoch": 0.70915907873594,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2996,
      "step": 2979
    },
    {
      "epoch": 0.7093971314646195,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 2980
    },
    {
      "epoch": 0.7096351841932989,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 2981
    },
    {
      "epoch": 0.7098732369219782,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 2982
    },
    {
      "epoch": 0.7101112896506576,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 2983
    },
    {
      "epoch": 0.710349342379337,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 2984
    },
    {
      "epoch": 0.7105873951080164,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 2985
    },
    {
      "epoch": 0.7108254478366959,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 2986
    },
    {
      "epoch": 0.7110635005653753,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 2987
    },
    {
      "epoch": 0.7113015532940546,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 2988
    },
    {
      "epoch": 0.711539606022734,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 2989
    },
    {
      "epoch": 0.7117776587514134,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 2990
    },
    {
      "epoch": 0.7120157114800928,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 2991
    },
    {
      "epoch": 0.7122537642087723,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.327,
      "step": 2992
    },
    {
      "epoch": 0.7124918169374517,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 2993
    },
    {
      "epoch": 0.712729869666131,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 2994
    },
    {
      "epoch": 0.7129679223948104,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4219,
      "step": 2995
    },
    {
      "epoch": 0.7132059751234898,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 2996
    },
    {
      "epoch": 0.7134440278521692,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 2997
    },
    {
      "epoch": 0.7136820805808487,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 2998
    },
    {
      "epoch": 0.7139201333095281,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4519,
      "step": 2999
    },
    {
      "epoch": 0.7141581860382075,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3985,
      "step": 3000
    },
    {
      "epoch": 0.7141581860382075,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 3000
    },
    {
      "epoch": 0.7143962387668868,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 3001
    },
    {
      "epoch": 0.7146342914955662,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 3002
    },
    {
      "epoch": 0.7148723442242456,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.7132,
      "step": 3003
    },
    {
      "epoch": 0.7151103969529251,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 3004
    },
    {
      "epoch": 0.7153484496816045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 3005
    },
    {
      "epoch": 0.7155865024102839,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 3006
    },
    {
      "epoch": 0.7158245551389633,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 3007
    },
    {
      "epoch": 0.7160626078676426,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.5476,
      "step": 3008
    },
    {
      "epoch": 0.7163006605963221,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 3009
    },
    {
      "epoch": 0.7165387133250015,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 3010
    },
    {
      "epoch": 0.7167767660536809,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 3011
    },
    {
      "epoch": 0.7170148187823603,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 3012
    },
    {
      "epoch": 0.7172528715110397,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 3013
    },
    {
      "epoch": 0.717490924239719,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 3014
    },
    {
      "epoch": 0.7177289769683985,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 3015
    },
    {
      "epoch": 0.7179670296970779,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 3016
    },
    {
      "epoch": 0.7182050824257573,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 3017
    },
    {
      "epoch": 0.7184431351544367,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 3018
    },
    {
      "epoch": 0.7186811878831161,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 3019
    },
    {
      "epoch": 0.7189192406117955,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 3020
    },
    {
      "epoch": 0.719157293340475,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 3021
    },
    {
      "epoch": 0.7193953460691543,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 3022
    },
    {
      "epoch": 0.7196333987978337,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 3023
    },
    {
      "epoch": 0.7198714515265131,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 3024
    },
    {
      "epoch": 0.7201095042551925,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4553,
      "step": 3025
    },
    {
      "epoch": 0.7203475569838719,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 3026
    },
    {
      "epoch": 0.7205856097125514,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2773,
      "step": 3027
    },
    {
      "epoch": 0.7208236624412308,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3533,
      "step": 3028
    },
    {
      "epoch": 0.7210617151699101,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.505,
      "step": 3029
    },
    {
      "epoch": 0.7212997678985895,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 3030
    },
    {
      "epoch": 0.7215378206272689,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 3031
    },
    {
      "epoch": 0.7217758733559484,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 3032
    },
    {
      "epoch": 0.7220139260846278,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2976,
      "step": 3033
    },
    {
      "epoch": 0.7222519788133072,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 3034
    },
    {
      "epoch": 0.7224900315419865,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 3035
    },
    {
      "epoch": 0.7227280842706659,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4136,
      "step": 3036
    },
    {
      "epoch": 0.7229661369993453,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 3037
    },
    {
      "epoch": 0.7232041897280248,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 3038
    },
    {
      "epoch": 0.7234422424567042,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 3039
    },
    {
      "epoch": 0.7236802951853836,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 3040
    },
    {
      "epoch": 0.723918347914063,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 3041
    },
    {
      "epoch": 0.7241564006427423,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 3042
    },
    {
      "epoch": 0.7243944533714217,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 3043
    },
    {
      "epoch": 0.7246325061001012,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 3044
    },
    {
      "epoch": 0.7248705588287806,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 3045
    },
    {
      "epoch": 0.72510861155746,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 3046
    },
    {
      "epoch": 0.7253466642861394,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.6344,
      "step": 3047
    },
    {
      "epoch": 0.7255847170148187,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 3048
    },
    {
      "epoch": 0.7258227697434981,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 3049
    },
    {
      "epoch": 0.7260608224721776,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 3050
    },
    {
      "epoch": 0.726298875200857,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 3051
    },
    {
      "epoch": 0.7265369279295364,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 3052
    },
    {
      "epoch": 0.7267749806582158,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3511,
      "step": 3053
    },
    {
      "epoch": 0.7270130333868952,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 3054
    },
    {
      "epoch": 0.7272510861155747,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3285,
      "step": 3055
    },
    {
      "epoch": 0.727489138844254,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 3056
    },
    {
      "epoch": 0.7277271915729334,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 3057
    },
    {
      "epoch": 0.7279652443016128,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 3058
    },
    {
      "epoch": 0.7282032970302922,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4834,
      "step": 3059
    },
    {
      "epoch": 0.7284413497589716,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 3060
    },
    {
      "epoch": 0.7286794024876511,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 3061
    },
    {
      "epoch": 0.7289174552163304,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 3062
    },
    {
      "epoch": 0.7291555079450098,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 3063
    },
    {
      "epoch": 0.7293935606736892,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3926,
      "step": 3064
    },
    {
      "epoch": 0.7296316134023686,
      "grad_norm": 7.5,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 3065
    },
    {
      "epoch": 0.729869666131048,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 3066
    },
    {
      "epoch": 0.7301077188597275,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4144,
      "step": 3067
    },
    {
      "epoch": 0.7303457715884069,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 3068
    },
    {
      "epoch": 0.7305838243170862,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 3069
    },
    {
      "epoch": 0.7308218770457656,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 3070
    },
    {
      "epoch": 0.731059929774445,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4144,
      "step": 3071
    },
    {
      "epoch": 0.7312979825031244,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 3072
    },
    {
      "epoch": 0.7315360352318039,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4498,
      "step": 3073
    },
    {
      "epoch": 0.7317740879604833,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.6152,
      "step": 3074
    },
    {
      "epoch": 0.7320121406891626,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 3075
    },
    {
      "epoch": 0.732250193417842,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 3076
    },
    {
      "epoch": 0.7324882461465214,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2235,
      "step": 3077
    },
    {
      "epoch": 0.7327262988752009,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3133,
      "step": 3078
    },
    {
      "epoch": 0.7329643516038803,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 3079
    },
    {
      "epoch": 0.7332024043325597,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 3080
    },
    {
      "epoch": 0.7334404570612391,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 3081
    },
    {
      "epoch": 0.7336785097899184,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 3082
    },
    {
      "epoch": 0.7339165625185978,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3716,
      "step": 3083
    },
    {
      "epoch": 0.7341546152472773,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2939,
      "step": 3084
    },
    {
      "epoch": 0.7343926679759567,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 3085
    },
    {
      "epoch": 0.7346307207046361,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3616,
      "step": 3086
    },
    {
      "epoch": 0.7348687734333155,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.5671,
      "step": 3087
    },
    {
      "epoch": 0.7351068261619949,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 3088
    },
    {
      "epoch": 0.7353448788906742,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4496,
      "step": 3089
    },
    {
      "epoch": 0.7355829316193537,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.6319,
      "step": 3090
    },
    {
      "epoch": 0.7358209843480331,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 3091
    },
    {
      "epoch": 0.7360590370767125,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 3092
    },
    {
      "epoch": 0.7362970898053919,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 3093
    },
    {
      "epoch": 0.7365351425340713,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 3094
    },
    {
      "epoch": 0.7367731952627506,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 3095
    },
    {
      "epoch": 0.7370112479914301,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 3096
    },
    {
      "epoch": 0.7372493007201095,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 3097
    },
    {
      "epoch": 0.7374873534487889,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3551,
      "step": 3098
    },
    {
      "epoch": 0.7377254061774683,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5529,
      "step": 3099
    },
    {
      "epoch": 0.7379634589061477,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4206,
      "step": 3100
    },
    {
      "epoch": 0.7379634589061477,
      "eval_runtime": 0.5112,
      "eval_samples_per_second": 1.956,
      "eval_steps_per_second": 1.956,
      "step": 3100
    },
    {
      "epoch": 0.7382015116348272,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4441,
      "step": 3101
    },
    {
      "epoch": 0.7384395643635066,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 3102
    },
    {
      "epoch": 0.7386776170921859,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5519,
      "step": 3103
    },
    {
      "epoch": 0.7389156698208653,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 3104
    },
    {
      "epoch": 0.7391537225495447,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 3105
    },
    {
      "epoch": 0.7393917752782241,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 3106
    },
    {
      "epoch": 0.7396298280069036,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 3107
    },
    {
      "epoch": 0.739867880735583,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 3108
    },
    {
      "epoch": 0.7401059334642623,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.7632,
      "step": 3109
    },
    {
      "epoch": 0.7403439861929417,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 3110
    },
    {
      "epoch": 0.7405820389216211,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 3111
    },
    {
      "epoch": 0.7408200916503005,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 3112
    },
    {
      "epoch": 0.74105814437898,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 3113
    },
    {
      "epoch": 0.7412961971076594,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 3114
    },
    {
      "epoch": 0.7415342498363388,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4621,
      "step": 3115
    },
    {
      "epoch": 0.7417723025650181,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 3116
    },
    {
      "epoch": 0.7420103552936975,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4161,
      "step": 3117
    },
    {
      "epoch": 0.7422484080223769,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 3118
    },
    {
      "epoch": 0.7424864607510564,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 3119
    },
    {
      "epoch": 0.7427245134797358,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 3120
    },
    {
      "epoch": 0.7429625662084152,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 3121
    },
    {
      "epoch": 0.7432006189370945,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 3122
    },
    {
      "epoch": 0.7434386716657739,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 3123
    },
    {
      "epoch": 0.7436767243944534,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 3124
    },
    {
      "epoch": 0.7439147771231328,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 3125
    },
    {
      "epoch": 0.7441528298518122,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3184,
      "step": 3126
    },
    {
      "epoch": 0.7443908825804916,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.425,
      "step": 3127
    },
    {
      "epoch": 0.744628935309171,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 3128
    },
    {
      "epoch": 0.7448669880378503,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 3129
    },
    {
      "epoch": 0.7451050407665298,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 3130
    },
    {
      "epoch": 0.7453430934952092,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 3131
    },
    {
      "epoch": 0.7455811462238886,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.431,
      "step": 3132
    },
    {
      "epoch": 0.745819198952568,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 3133
    },
    {
      "epoch": 0.7460572516812474,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 3134
    },
    {
      "epoch": 0.7462953044099268,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4056,
      "step": 3135
    },
    {
      "epoch": 0.7465333571386062,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4294,
      "step": 3136
    },
    {
      "epoch": 0.7467714098672856,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2892,
      "step": 3137
    },
    {
      "epoch": 0.747009462595965,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4723,
      "step": 3138
    },
    {
      "epoch": 0.7472475153246444,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 3139
    },
    {
      "epoch": 0.7474855680533238,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 3140
    },
    {
      "epoch": 0.7477236207820032,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 3141
    },
    {
      "epoch": 0.7479616735106827,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4202,
      "step": 3142
    },
    {
      "epoch": 0.748199726239362,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 3143
    },
    {
      "epoch": 0.7484377789680414,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 3144
    },
    {
      "epoch": 0.7486758316967208,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 3145
    },
    {
      "epoch": 0.7489138844254002,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3172,
      "step": 3146
    },
    {
      "epoch": 0.7491519371540797,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 3147
    },
    {
      "epoch": 0.7493899898827591,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3119,
      "step": 3148
    },
    {
      "epoch": 0.7496280426114384,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 3149
    },
    {
      "epoch": 0.7498660953401178,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 3150
    },
    {
      "epoch": 0.7501041480687972,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 3151
    },
    {
      "epoch": 0.7503422007974766,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 3152
    },
    {
      "epoch": 0.7505802535261561,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 3153
    },
    {
      "epoch": 0.7508183062548355,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 3154
    },
    {
      "epoch": 0.7510563589835149,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2937,
      "step": 3155
    },
    {
      "epoch": 0.7512944117121942,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 3156
    },
    {
      "epoch": 0.7515324644408736,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4003,
      "step": 3157
    },
    {
      "epoch": 0.751770517169553,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 3158
    },
    {
      "epoch": 0.7520085698982325,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3311,
      "step": 3159
    },
    {
      "epoch": 0.7522466226269119,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 3160
    },
    {
      "epoch": 0.7524846753555913,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.6124,
      "step": 3161
    },
    {
      "epoch": 0.7527227280842707,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 3162
    },
    {
      "epoch": 0.75296078081295,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 3163
    },
    {
      "epoch": 0.7531988335416294,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 3164
    },
    {
      "epoch": 0.7534368862703089,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 3165
    },
    {
      "epoch": 0.7536749389989883,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 3166
    },
    {
      "epoch": 0.7539129917276677,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 3167
    },
    {
      "epoch": 0.7541510444563471,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 3168
    },
    {
      "epoch": 0.7543890971850264,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 3169
    },
    {
      "epoch": 0.7546271499137059,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 3170
    },
    {
      "epoch": 0.7548652026423853,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 3171
    },
    {
      "epoch": 0.7551032553710647,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 3172
    },
    {
      "epoch": 0.7553413080997441,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 3173
    },
    {
      "epoch": 0.7555793608284235,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.4035,
      "step": 3174
    },
    {
      "epoch": 0.7558174135571029,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 3175
    },
    {
      "epoch": 0.7560554662857824,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 3176
    },
    {
      "epoch": 0.7562935190144617,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 3177
    },
    {
      "epoch": 0.7565315717431411,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 3178
    },
    {
      "epoch": 0.7567696244718205,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 3179
    },
    {
      "epoch": 0.7570076772004999,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 3180
    },
    {
      "epoch": 0.7572457299291793,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4979,
      "step": 3181
    },
    {
      "epoch": 0.7574837826578588,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 3182
    },
    {
      "epoch": 0.7577218353865381,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 3183
    },
    {
      "epoch": 0.7579598881152175,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 3184
    },
    {
      "epoch": 0.7581979408438969,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 3185
    },
    {
      "epoch": 0.7584359935725763,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 3186
    },
    {
      "epoch": 0.7586740463012557,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 3187
    },
    {
      "epoch": 0.7589120990299352,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3301,
      "step": 3188
    },
    {
      "epoch": 0.7591501517586146,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3674,
      "step": 3189
    },
    {
      "epoch": 0.7593882044872939,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3634,
      "step": 3190
    },
    {
      "epoch": 0.7596262572159733,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 3191
    },
    {
      "epoch": 0.7598643099446527,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 3192
    },
    {
      "epoch": 0.7601023626733321,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 3193
    },
    {
      "epoch": 0.7603404154020116,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 3194
    },
    {
      "epoch": 0.760578468130691,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 3195
    },
    {
      "epoch": 0.7608165208593703,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 3196
    },
    {
      "epoch": 0.7610545735880497,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 3197
    },
    {
      "epoch": 0.7612926263167291,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 3198
    },
    {
      "epoch": 0.7615306790454086,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 3199
    },
    {
      "epoch": 0.761768731774088,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 3200
    },
    {
      "epoch": 0.761768731774088,
      "eval_runtime": 0.5155,
      "eval_samples_per_second": 1.94,
      "eval_steps_per_second": 1.94,
      "step": 3200
    },
    {
      "epoch": 0.7620067845027674,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 3201
    },
    {
      "epoch": 0.7622448372314468,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2748,
      "step": 3202
    },
    {
      "epoch": 0.7624828899601261,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 3203
    },
    {
      "epoch": 0.7627209426888055,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2922,
      "step": 3204
    },
    {
      "epoch": 0.762958995417485,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3756,
      "step": 3205
    },
    {
      "epoch": 0.7631970481461644,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 3206
    },
    {
      "epoch": 0.7634351008748438,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 3207
    },
    {
      "epoch": 0.7636731536035232,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 3208
    },
    {
      "epoch": 0.7639112063322026,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.278,
      "step": 3209
    },
    {
      "epoch": 0.7641492590608819,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 3210
    },
    {
      "epoch": 0.7643873117895614,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4639,
      "step": 3211
    },
    {
      "epoch": 0.7646253645182408,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 3212
    },
    {
      "epoch": 0.7648634172469202,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 3213
    },
    {
      "epoch": 0.7651014699755996,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 3214
    },
    {
      "epoch": 0.765339522704279,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 3215
    },
    {
      "epoch": 0.7655775754329583,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 3216
    },
    {
      "epoch": 0.7658156281616378,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4545,
      "step": 3217
    },
    {
      "epoch": 0.7660536808903172,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 3218
    },
    {
      "epoch": 0.7662917336189966,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 3219
    },
    {
      "epoch": 0.766529786347676,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 3220
    },
    {
      "epoch": 0.7667678390763554,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.4337,
      "step": 3221
    },
    {
      "epoch": 0.7670058918050349,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 3222
    },
    {
      "epoch": 0.7672439445337143,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 3223
    },
    {
      "epoch": 0.7674819972623936,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 3224
    },
    {
      "epoch": 0.767720049991073,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 3225
    },
    {
      "epoch": 0.7679581027197524,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 3226
    },
    {
      "epoch": 0.7681961554484318,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 3227
    },
    {
      "epoch": 0.7684342081771113,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.237,
      "step": 3228
    },
    {
      "epoch": 0.7686722609057907,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 3229
    },
    {
      "epoch": 0.76891031363447,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 3230
    },
    {
      "epoch": 0.7691483663631494,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 3231
    },
    {
      "epoch": 0.7693864190918288,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 3232
    },
    {
      "epoch": 0.7696244718205082,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 3233
    },
    {
      "epoch": 0.7698625245491877,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4395,
      "step": 3234
    },
    {
      "epoch": 0.7701005772778671,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 3235
    },
    {
      "epoch": 0.7703386300065465,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.4056,
      "step": 3236
    },
    {
      "epoch": 0.7705766827352258,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4695,
      "step": 3237
    },
    {
      "epoch": 0.7708147354639052,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.4393,
      "step": 3238
    },
    {
      "epoch": 0.7710527881925846,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 3239
    },
    {
      "epoch": 0.7712908409212641,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 3240
    },
    {
      "epoch": 0.7715288936499435,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 3241
    },
    {
      "epoch": 0.7717669463786229,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 3242
    },
    {
      "epoch": 0.7720049991073022,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 3243
    },
    {
      "epoch": 0.7722430518359816,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 3244
    },
    {
      "epoch": 0.7724811045646611,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 3245
    },
    {
      "epoch": 0.7727191572933405,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 3246
    },
    {
      "epoch": 0.7729572100220199,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 3247
    },
    {
      "epoch": 0.7731952627506993,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 3248
    },
    {
      "epoch": 0.7734333154793787,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 3249
    },
    {
      "epoch": 0.773671368208058,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 3250
    },
    {
      "epoch": 0.7739094209367375,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 3251
    },
    {
      "epoch": 0.7741474736654169,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 3252
    },
    {
      "epoch": 0.7743855263940963,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 3253
    },
    {
      "epoch": 0.7746235791227757,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 3254
    },
    {
      "epoch": 0.7748616318514551,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 3255
    },
    {
      "epoch": 0.7750996845801345,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 3256
    },
    {
      "epoch": 0.775337737308814,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 3257
    },
    {
      "epoch": 0.7755757900374933,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2958,
      "step": 3258
    },
    {
      "epoch": 0.7758138427661727,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 3259
    },
    {
      "epoch": 0.7760518954948521,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 3260
    },
    {
      "epoch": 0.7762899482235315,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3224,
      "step": 3261
    },
    {
      "epoch": 0.7765280009522109,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4254,
      "step": 3262
    },
    {
      "epoch": 0.7767660536808904,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 3263
    },
    {
      "epoch": 0.7770041064095697,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 3264
    },
    {
      "epoch": 0.7772421591382491,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 3265
    },
    {
      "epoch": 0.7774802118669285,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.395,
      "step": 3266
    },
    {
      "epoch": 0.7777182645956079,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 3267
    },
    {
      "epoch": 0.7779563173242874,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4507,
      "step": 3268
    },
    {
      "epoch": 0.7781943700529668,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 3269
    },
    {
      "epoch": 0.7784324227816461,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5094,
      "step": 3270
    },
    {
      "epoch": 0.7786704755103255,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 3271
    },
    {
      "epoch": 0.7789085282390049,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 3272
    },
    {
      "epoch": 0.7791465809676843,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 3273
    },
    {
      "epoch": 0.7793846336963638,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 3274
    },
    {
      "epoch": 0.7796226864250432,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3977,
      "step": 3275
    },
    {
      "epoch": 0.7798607391537226,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3677,
      "step": 3276
    },
    {
      "epoch": 0.7800987918824019,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 3277
    },
    {
      "epoch": 0.7803368446110813,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 3278
    },
    {
      "epoch": 0.7805748973397607,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3281,
      "step": 3279
    },
    {
      "epoch": 0.7808129500684402,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3537,
      "step": 3280
    },
    {
      "epoch": 0.7810510027971196,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 3281
    },
    {
      "epoch": 0.781289055525799,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 3282
    },
    {
      "epoch": 0.7815271082544784,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 3283
    },
    {
      "epoch": 0.7817651609831577,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3142,
      "step": 3284
    },
    {
      "epoch": 0.7820032137118371,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 3285
    },
    {
      "epoch": 0.7822412664405166,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 3286
    },
    {
      "epoch": 0.782479319169196,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 3287
    },
    {
      "epoch": 0.7827173718978754,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 3288
    },
    {
      "epoch": 0.7829554246265548,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 3289
    },
    {
      "epoch": 0.7831934773552341,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 3290
    },
    {
      "epoch": 0.7834315300839136,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 3291
    },
    {
      "epoch": 0.783669582812593,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 3292
    },
    {
      "epoch": 0.7839076355412724,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 3293
    },
    {
      "epoch": 0.7841456882699518,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.412,
      "step": 3294
    },
    {
      "epoch": 0.7843837409986312,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.501,
      "step": 3295
    },
    {
      "epoch": 0.7846217937273106,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 3296
    },
    {
      "epoch": 0.78485984645599,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2881,
      "step": 3297
    },
    {
      "epoch": 0.7850978991846694,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 3298
    },
    {
      "epoch": 0.7853359519133488,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 3299
    },
    {
      "epoch": 0.7855740046420282,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 3300
    },
    {
      "epoch": 0.7855740046420282,
      "eval_runtime": 0.5051,
      "eval_samples_per_second": 1.98,
      "eval_steps_per_second": 1.98,
      "step": 3300
    },
    {
      "epoch": 0.7858120573707076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 3301
    },
    {
      "epoch": 0.786050110099387,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 3302
    },
    {
      "epoch": 0.7862881628280665,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 3303
    },
    {
      "epoch": 0.7865262155567458,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 3304
    },
    {
      "epoch": 0.7867642682854252,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.278,
      "step": 3305
    },
    {
      "epoch": 0.7870023210141046,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 3306
    },
    {
      "epoch": 0.787240373742784,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 3307
    },
    {
      "epoch": 0.7874784264714634,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 3308
    },
    {
      "epoch": 0.7877164792001429,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3816,
      "step": 3309
    },
    {
      "epoch": 0.7879545319288223,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 3310
    },
    {
      "epoch": 0.7881925846575016,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 3311
    },
    {
      "epoch": 0.788430637386181,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4547,
      "step": 3312
    },
    {
      "epoch": 0.7886686901148604,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 3313
    },
    {
      "epoch": 0.7889067428435399,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 3314
    },
    {
      "epoch": 0.7891447955722193,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 3315
    },
    {
      "epoch": 0.7893828483008987,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 3316
    },
    {
      "epoch": 0.789620901029578,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 3317
    },
    {
      "epoch": 0.7898589537582574,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.4744,
      "step": 3318
    },
    {
      "epoch": 0.7900970064869368,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 3319
    },
    {
      "epoch": 0.7903350592156163,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4886,
      "step": 3320
    },
    {
      "epoch": 0.7905731119442957,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 3321
    },
    {
      "epoch": 0.7908111646729751,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 3322
    },
    {
      "epoch": 0.7910492174016545,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3661,
      "step": 3323
    },
    {
      "epoch": 0.7912872701303338,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2744,
      "step": 3324
    },
    {
      "epoch": 0.7915253228590132,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 3325
    },
    {
      "epoch": 0.7917633755876927,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.189,
      "step": 3326
    },
    {
      "epoch": 0.7920014283163721,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 3327
    },
    {
      "epoch": 0.7922394810450515,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 3328
    },
    {
      "epoch": 0.7924775337737309,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 3329
    },
    {
      "epoch": 0.7927155865024103,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 3330
    },
    {
      "epoch": 0.7929536392310896,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 3331
    },
    {
      "epoch": 0.7931916919597691,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 3332
    },
    {
      "epoch": 0.7934297446884485,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.442,
      "step": 3333
    },
    {
      "epoch": 0.7936677974171279,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 3334
    },
    {
      "epoch": 0.7939058501458073,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 3335
    },
    {
      "epoch": 0.7941439028744867,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 3336
    },
    {
      "epoch": 0.7943819556031662,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 3337
    },
    {
      "epoch": 0.7946200083318455,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 3338
    },
    {
      "epoch": 0.7948580610605249,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 3339
    },
    {
      "epoch": 0.7950961137892043,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 3340
    },
    {
      "epoch": 0.7953341665178837,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.3392,
      "step": 3341
    },
    {
      "epoch": 0.7955722192465631,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 3342
    },
    {
      "epoch": 0.7958102719752426,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3935,
      "step": 3343
    },
    {
      "epoch": 0.796048324703922,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 3344
    },
    {
      "epoch": 0.7962863774326013,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 3345
    },
    {
      "epoch": 0.7965244301612807,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 3346
    },
    {
      "epoch": 0.7967624828899601,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 3347
    },
    {
      "epoch": 0.7970005356186395,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 3348
    },
    {
      "epoch": 0.797238588347319,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 3349
    },
    {
      "epoch": 0.7974766410759984,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 3350
    },
    {
      "epoch": 0.7977146938046777,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3217,
      "step": 3351
    },
    {
      "epoch": 0.7979527465333571,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 3352
    },
    {
      "epoch": 0.7981907992620365,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 3353
    },
    {
      "epoch": 0.7984288519907159,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 3354
    },
    {
      "epoch": 0.7986669047193954,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 3355
    },
    {
      "epoch": 0.7989049574480748,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 3356
    },
    {
      "epoch": 0.7991430101767542,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 3357
    },
    {
      "epoch": 0.7993810629054335,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 3358
    },
    {
      "epoch": 0.7996191156341129,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 3359
    },
    {
      "epoch": 0.7998571683627923,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 3360
    },
    {
      "epoch": 0.8000952210914718,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4979,
      "step": 3361
    },
    {
      "epoch": 0.8003332738201512,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 3362
    },
    {
      "epoch": 0.8005713265488306,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 3363
    },
    {
      "epoch": 0.80080937927751,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 3364
    },
    {
      "epoch": 0.8010474320061893,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 3365
    },
    {
      "epoch": 0.8012854847348688,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 3366
    },
    {
      "epoch": 0.8015235374635482,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 3367
    },
    {
      "epoch": 0.8017615901922276,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 3368
    },
    {
      "epoch": 0.801999642920907,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.4346,
      "step": 3369
    },
    {
      "epoch": 0.8022376956495864,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 3370
    },
    {
      "epoch": 0.8024757483782657,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 3371
    },
    {
      "epoch": 0.8027138011069452,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.8306,
      "step": 3372
    },
    {
      "epoch": 0.8029518538356246,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5078,
      "step": 3373
    },
    {
      "epoch": 0.803189906564304,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 3374
    },
    {
      "epoch": 0.8034279592929834,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 3375
    },
    {
      "epoch": 0.8036660120216628,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 3376
    },
    {
      "epoch": 0.8039040647503422,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 3377
    },
    {
      "epoch": 0.8041421174790216,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 3378
    },
    {
      "epoch": 0.804380170207701,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 3379
    },
    {
      "epoch": 0.8046182229363804,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 3380
    },
    {
      "epoch": 0.8048562756650598,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 3381
    },
    {
      "epoch": 0.8050943283937392,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 3382
    },
    {
      "epoch": 0.8053323811224186,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 3383
    },
    {
      "epoch": 0.8055704338510981,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.6642,
      "step": 3384
    },
    {
      "epoch": 0.8058084865797774,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 3385
    },
    {
      "epoch": 0.8060465393084568,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 3386
    },
    {
      "epoch": 0.8062845920371362,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 3387
    },
    {
      "epoch": 0.8065226447658156,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 3388
    },
    {
      "epoch": 0.8067606974944951,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 3389
    },
    {
      "epoch": 0.8069987502231745,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.6041,
      "step": 3390
    },
    {
      "epoch": 0.8072368029518538,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 3391
    },
    {
      "epoch": 0.8074748556805332,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 3392
    },
    {
      "epoch": 0.8077129084092126,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 3393
    },
    {
      "epoch": 0.807950961137892,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 3394
    },
    {
      "epoch": 0.8081890138665715,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2728,
      "step": 3395
    },
    {
      "epoch": 0.8084270665952509,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 3396
    },
    {
      "epoch": 0.8086651193239303,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 3397
    },
    {
      "epoch": 0.8089031720526096,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3477,
      "step": 3398
    },
    {
      "epoch": 0.809141224781289,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 3399
    },
    {
      "epoch": 0.8093792775099684,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 3400
    },
    {
      "epoch": 0.8093792775099684,
      "eval_runtime": 0.5089,
      "eval_samples_per_second": 1.965,
      "eval_steps_per_second": 1.965,
      "step": 3400
    },
    {
      "epoch": 0.8096173302386479,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 3401
    },
    {
      "epoch": 0.8098553829673273,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 3402
    },
    {
      "epoch": 0.8100934356960067,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 3403
    },
    {
      "epoch": 0.810331488424686,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 3404
    },
    {
      "epoch": 0.8105695411533654,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 3405
    },
    {
      "epoch": 0.8108075938820448,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 3406
    },
    {
      "epoch": 0.8110456466107243,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 3407
    },
    {
      "epoch": 0.8112836993394037,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 3408
    },
    {
      "epoch": 0.8115217520680831,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 3409
    },
    {
      "epoch": 0.8117598047967625,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.5066,
      "step": 3410
    },
    {
      "epoch": 0.8119978575254418,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 3411
    },
    {
      "epoch": 0.8122359102541213,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 3412
    },
    {
      "epoch": 0.8124739629828007,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2791,
      "step": 3413
    },
    {
      "epoch": 0.8127120157114801,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 3414
    },
    {
      "epoch": 0.8129500684401595,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 3415
    },
    {
      "epoch": 0.8131881211688389,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 3416
    },
    {
      "epoch": 0.8134261738975183,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 3417
    },
    {
      "epoch": 0.8136642266261978,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 3418
    },
    {
      "epoch": 0.8139022793548771,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2542,
      "step": 3419
    },
    {
      "epoch": 0.8141403320835565,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4662,
      "step": 3420
    },
    {
      "epoch": 0.8143783848122359,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 3421
    },
    {
      "epoch": 0.8146164375409153,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 3422
    },
    {
      "epoch": 0.8148544902695947,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 3423
    },
    {
      "epoch": 0.8150925429982742,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 3424
    },
    {
      "epoch": 0.8153305957269535,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 3425
    },
    {
      "epoch": 0.8155686484556329,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 3426
    },
    {
      "epoch": 0.8158067011843123,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2757,
      "step": 3427
    },
    {
      "epoch": 0.8160447539129917,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 3428
    },
    {
      "epoch": 0.8162828066416711,
      "grad_norm": 2.578125,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 3429
    },
    {
      "epoch": 0.8165208593703506,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 3430
    },
    {
      "epoch": 0.81675891209903,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 3431
    },
    {
      "epoch": 0.8169969648277093,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2643,
      "step": 3432
    },
    {
      "epoch": 0.8172350175563887,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 3433
    },
    {
      "epoch": 0.8174730702850681,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 3434
    },
    {
      "epoch": 0.8177111230137476,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 3435
    },
    {
      "epoch": 0.817949175742427,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 3436
    },
    {
      "epoch": 0.8181872284711064,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4429,
      "step": 3437
    },
    {
      "epoch": 0.8184252811997857,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 3438
    },
    {
      "epoch": 0.8186633339284651,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 3439
    },
    {
      "epoch": 0.8189013866571445,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 3440
    },
    {
      "epoch": 0.819139439385824,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3248,
      "step": 3441
    },
    {
      "epoch": 0.8193774921145034,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3578,
      "step": 3442
    },
    {
      "epoch": 0.8196155448431828,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.3723,
      "step": 3443
    },
    {
      "epoch": 0.8198535975718622,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 3444
    },
    {
      "epoch": 0.8200916503005415,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 3445
    },
    {
      "epoch": 0.8203297030292209,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 3446
    },
    {
      "epoch": 0.8205677557579004,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4285,
      "step": 3447
    },
    {
      "epoch": 0.8208058084865798,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.456,
      "step": 3448
    },
    {
      "epoch": 0.8210438612152592,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 3449
    },
    {
      "epoch": 0.8212819139439386,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 3450
    },
    {
      "epoch": 0.821519966672618,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3349,
      "step": 3451
    },
    {
      "epoch": 0.8217580194012973,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 3452
    },
    {
      "epoch": 0.8219960721299768,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 3453
    },
    {
      "epoch": 0.8222341248586562,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 3454
    },
    {
      "epoch": 0.8224721775873356,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 3455
    },
    {
      "epoch": 0.822710230316015,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 3456
    },
    {
      "epoch": 0.8229482830446944,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 3457
    },
    {
      "epoch": 0.8231863357733739,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 3458
    },
    {
      "epoch": 0.8234243885020532,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 3459
    },
    {
      "epoch": 0.8236624412307326,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 3460
    },
    {
      "epoch": 0.823900493959412,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.8007,
      "step": 3461
    },
    {
      "epoch": 0.8241385466880914,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 3462
    },
    {
      "epoch": 0.8243765994167708,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 3463
    },
    {
      "epoch": 0.8246146521454503,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 3464
    },
    {
      "epoch": 0.8248527048741296,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3421,
      "step": 3465
    },
    {
      "epoch": 0.825090757602809,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 2.3428,
      "step": 3466
    },
    {
      "epoch": 0.8253288103314884,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 3467
    },
    {
      "epoch": 0.8255668630601678,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 3468
    },
    {
      "epoch": 0.8258049157888472,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 3469
    },
    {
      "epoch": 0.8260429685175267,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 3470
    },
    {
      "epoch": 0.8262810212462061,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 3471
    },
    {
      "epoch": 0.8265190739748854,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 3472
    },
    {
      "epoch": 0.8267571267035648,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 3473
    },
    {
      "epoch": 0.8269951794322442,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 3474
    },
    {
      "epoch": 0.8272332321609236,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 3475
    },
    {
      "epoch": 0.8274712848896031,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 3476
    },
    {
      "epoch": 0.8277093376182825,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 3477
    },
    {
      "epoch": 0.8279473903469619,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 3478
    },
    {
      "epoch": 0.8281854430756412,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2916,
      "step": 3479
    },
    {
      "epoch": 0.8284234958043206,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 3480
    },
    {
      "epoch": 0.8286615485330001,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3928,
      "step": 3481
    },
    {
      "epoch": 0.8288996012616795,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 3482
    },
    {
      "epoch": 0.8291376539903589,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 3483
    },
    {
      "epoch": 0.8293757067190383,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 3484
    },
    {
      "epoch": 0.8296137594477176,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 3485
    },
    {
      "epoch": 0.829851812176397,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 3486
    },
    {
      "epoch": 0.8300898649050765,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 3487
    },
    {
      "epoch": 0.8303279176337559,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 3488
    },
    {
      "epoch": 0.8305659703624353,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 3489
    },
    {
      "epoch": 0.8308040230911147,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 3490
    },
    {
      "epoch": 0.8310420758197941,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 3491
    },
    {
      "epoch": 0.8312801285484734,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4259,
      "step": 3492
    },
    {
      "epoch": 0.8315181812771529,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 3493
    },
    {
      "epoch": 0.8317562340058323,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3608,
      "step": 3494
    },
    {
      "epoch": 0.8319942867345117,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 3495
    },
    {
      "epoch": 0.8322323394631911,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 3496
    },
    {
      "epoch": 0.8324703921918705,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 3497
    },
    {
      "epoch": 0.8327084449205499,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 3498
    },
    {
      "epoch": 0.8329464976492293,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4738,
      "step": 3499
    },
    {
      "epoch": 0.8331845503779087,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 3500
    },
    {
      "epoch": 0.8331845503779087,
      "eval_runtime": 0.5206,
      "eval_samples_per_second": 1.921,
      "eval_steps_per_second": 1.921,
      "step": 3500
    },
    {
      "epoch": 0.8334226031065881,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 3501
    },
    {
      "epoch": 0.8336606558352675,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2674,
      "step": 3502
    },
    {
      "epoch": 0.8338987085639469,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 3503
    },
    {
      "epoch": 0.8341367612926264,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 3504
    },
    {
      "epoch": 0.8343748140213058,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 3505
    },
    {
      "epoch": 0.8346128667499851,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 3506
    },
    {
      "epoch": 0.8348509194786645,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 3507
    },
    {
      "epoch": 0.8350889722073439,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 3508
    },
    {
      "epoch": 0.8353270249360233,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 3509
    },
    {
      "epoch": 0.8355650776647028,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.4299,
      "step": 3510
    },
    {
      "epoch": 0.8358031303933822,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2916,
      "step": 3511
    },
    {
      "epoch": 0.8360411831220615,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3801,
      "step": 3512
    },
    {
      "epoch": 0.8362792358507409,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 3513
    },
    {
      "epoch": 0.8365172885794203,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.6587,
      "step": 3514
    },
    {
      "epoch": 0.8367553413080997,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 3515
    },
    {
      "epoch": 0.8369933940367792,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3257,
      "step": 3516
    },
    {
      "epoch": 0.8372314467654586,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.3553,
      "step": 3517
    },
    {
      "epoch": 0.837469499494138,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 3518
    },
    {
      "epoch": 0.8377075522228173,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 3519
    },
    {
      "epoch": 0.8379456049514967,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 3520
    },
    {
      "epoch": 0.8381836576801761,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 3521
    },
    {
      "epoch": 0.8384217104088556,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4523,
      "step": 3522
    },
    {
      "epoch": 0.838659763137535,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 3523
    },
    {
      "epoch": 0.8388978158662144,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 3524
    },
    {
      "epoch": 0.8391358685948938,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 3525
    },
    {
      "epoch": 0.8393739213235731,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 3526
    },
    {
      "epoch": 0.8396119740522525,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 3527
    },
    {
      "epoch": 0.839850026780932,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 3528
    },
    {
      "epoch": 0.8400880795096114,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 3529
    },
    {
      "epoch": 0.8403261322382908,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 3530
    },
    {
      "epoch": 0.8405641849669702,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 3531
    },
    {
      "epoch": 0.8408022376956495,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 3532
    },
    {
      "epoch": 0.841040290424329,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 3533
    },
    {
      "epoch": 0.8412783431530084,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 3534
    },
    {
      "epoch": 0.8415163958816878,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 3535
    },
    {
      "epoch": 0.8417544486103672,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 3536
    },
    {
      "epoch": 0.8419925013390466,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2796,
      "step": 3537
    },
    {
      "epoch": 0.842230554067726,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 3538
    },
    {
      "epoch": 0.8424686067964055,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 3539
    },
    {
      "epoch": 0.8427066595250848,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 3540
    },
    {
      "epoch": 0.8429447122537642,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 3541
    },
    {
      "epoch": 0.8431827649824436,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 3542
    },
    {
      "epoch": 0.843420817711123,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 3543
    },
    {
      "epoch": 0.8436588704398024,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 3544
    },
    {
      "epoch": 0.8438969231684819,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4653,
      "step": 3545
    },
    {
      "epoch": 0.8441349758971612,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.5018,
      "step": 3546
    },
    {
      "epoch": 0.8443730286258406,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 3547
    },
    {
      "epoch": 0.84461108135452,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 3548
    },
    {
      "epoch": 0.8448491340831994,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 3549
    },
    {
      "epoch": 0.8450871868118788,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3781,
      "step": 3550
    },
    {
      "epoch": 0.8453252395405583,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 3551
    },
    {
      "epoch": 0.8455632922692377,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2244,
      "step": 3552
    },
    {
      "epoch": 0.845801344997917,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 3553
    },
    {
      "epoch": 0.8460393977265964,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 3554
    },
    {
      "epoch": 0.8462774504552758,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 3555
    },
    {
      "epoch": 0.8465155031839553,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 3556
    },
    {
      "epoch": 0.8467535559126347,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 3557
    },
    {
      "epoch": 0.8469916086413141,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3664,
      "step": 3558
    },
    {
      "epoch": 0.8472296613699934,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 3559
    },
    {
      "epoch": 0.8474677140986728,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 3560
    },
    {
      "epoch": 0.8477057668273522,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 3561
    },
    {
      "epoch": 0.8479438195560317,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 3562
    },
    {
      "epoch": 0.8481818722847111,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 3563
    },
    {
      "epoch": 0.8484199250133905,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 3564
    },
    {
      "epoch": 0.8486579777420699,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 3565
    },
    {
      "epoch": 0.8488960304707492,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 3566
    },
    {
      "epoch": 0.8491340831994286,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 3567
    },
    {
      "epoch": 0.8493721359281081,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 3568
    },
    {
      "epoch": 0.8496101886567875,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 3569
    },
    {
      "epoch": 0.8498482413854669,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 3570
    },
    {
      "epoch": 0.8500862941141463,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 3571
    },
    {
      "epoch": 0.8503243468428257,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 3572
    },
    {
      "epoch": 0.850562399571505,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 3573
    },
    {
      "epoch": 0.8508004523001845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.7048,
      "step": 3574
    },
    {
      "epoch": 0.8510385050288639,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4082,
      "step": 3575
    },
    {
      "epoch": 0.8512765577575433,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 3576
    },
    {
      "epoch": 0.8515146104862227,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.5164,
      "step": 3577
    },
    {
      "epoch": 0.8517526632149021,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 3578
    },
    {
      "epoch": 0.8519907159435816,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 3579
    },
    {
      "epoch": 0.8522287686722609,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4437,
      "step": 3580
    },
    {
      "epoch": 0.8524668214009403,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3424,
      "step": 3581
    },
    {
      "epoch": 0.8527048741296197,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 3582
    },
    {
      "epoch": 0.8529429268582991,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 3583
    },
    {
      "epoch": 0.8531809795869785,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 3584
    },
    {
      "epoch": 0.853419032315658,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 3585
    },
    {
      "epoch": 0.8536570850443373,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 3586
    },
    {
      "epoch": 0.8538951377730167,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 3587
    },
    {
      "epoch": 0.8541331905016961,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 3588
    },
    {
      "epoch": 0.8543712432303755,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.5279,
      "step": 3589
    },
    {
      "epoch": 0.8546092959590549,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 3590
    },
    {
      "epoch": 0.8548473486877344,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.6236,
      "step": 3591
    },
    {
      "epoch": 0.8550854014164138,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 3592
    },
    {
      "epoch": 0.8553234541450931,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 3593
    },
    {
      "epoch": 0.8555615068737725,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.5018,
      "step": 3594
    },
    {
      "epoch": 0.8557995596024519,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3588,
      "step": 3595
    },
    {
      "epoch": 0.8560376123311313,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3566,
      "step": 3596
    },
    {
      "epoch": 0.8562756650598108,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 3597
    },
    {
      "epoch": 0.8565137177884902,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 3598
    },
    {
      "epoch": 0.8567517705171696,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.462,
      "step": 3599
    },
    {
      "epoch": 0.8569898232458489,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4532,
      "step": 3600
    },
    {
      "epoch": 0.8569898232458489,
      "eval_runtime": 0.5285,
      "eval_samples_per_second": 1.892,
      "eval_steps_per_second": 1.892,
      "step": 3600
    },
    {
      "epoch": 0.8572278759745283,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 3601
    },
    {
      "epoch": 0.8574659287032078,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 3602
    },
    {
      "epoch": 0.8577039814318872,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 3603
    },
    {
      "epoch": 0.8579420341605666,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4073,
      "step": 3604
    },
    {
      "epoch": 0.858180086889246,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3629,
      "step": 3605
    },
    {
      "epoch": 0.8584181396179253,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2643,
      "step": 3606
    },
    {
      "epoch": 0.8586561923466047,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 3607
    },
    {
      "epoch": 0.8588942450752842,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 3608
    },
    {
      "epoch": 0.8591322978039636,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 3609
    },
    {
      "epoch": 0.859370350532643,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 3610
    },
    {
      "epoch": 0.8596084032613224,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 3611
    },
    {
      "epoch": 0.8598464559900018,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3576,
      "step": 3612
    },
    {
      "epoch": 0.8600845087186811,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 3613
    },
    {
      "epoch": 0.8603225614473606,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 3614
    },
    {
      "epoch": 0.86056061417604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 3615
    },
    {
      "epoch": 0.8607986669047194,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 3616
    },
    {
      "epoch": 0.8610367196333988,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 3617
    },
    {
      "epoch": 0.8612747723620782,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 3618
    },
    {
      "epoch": 0.8615128250907576,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 3619
    },
    {
      "epoch": 0.861750877819437,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 3620
    },
    {
      "epoch": 0.8619889305481164,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 3621
    },
    {
      "epoch": 0.8622269832767958,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 3622
    },
    {
      "epoch": 0.8624650360054752,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 3623
    },
    {
      "epoch": 0.8627030887341546,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 3624
    },
    {
      "epoch": 0.8629411414628341,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 3625
    },
    {
      "epoch": 0.8631791941915135,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4424,
      "step": 3626
    },
    {
      "epoch": 0.8634172469201928,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3774,
      "step": 3627
    },
    {
      "epoch": 0.8636552996488722,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 3628
    },
    {
      "epoch": 0.8638933523775516,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 3629
    },
    {
      "epoch": 0.864131405106231,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 3630
    },
    {
      "epoch": 0.8643694578349105,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 3631
    },
    {
      "epoch": 0.8646075105635899,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 3632
    },
    {
      "epoch": 0.8648455632922692,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3268,
      "step": 3633
    },
    {
      "epoch": 0.8650836160209486,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 3634
    },
    {
      "epoch": 0.865321668749628,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 3635
    },
    {
      "epoch": 0.8655597214783074,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 3636
    },
    {
      "epoch": 0.8657977742069869,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3003,
      "step": 3637
    },
    {
      "epoch": 0.8660358269356663,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 3638
    },
    {
      "epoch": 0.8662738796643457,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 3639
    },
    {
      "epoch": 0.866511932393025,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 3640
    },
    {
      "epoch": 0.8667499851217044,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 3641
    },
    {
      "epoch": 0.8669880378503838,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4682,
      "step": 3642
    },
    {
      "epoch": 0.8672260905790633,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3509,
      "step": 3643
    },
    {
      "epoch": 0.8674641433077427,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 3644
    },
    {
      "epoch": 0.8677021960364221,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 3645
    },
    {
      "epoch": 0.8679402487651015,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 3646
    },
    {
      "epoch": 0.8681783014937808,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.362,
      "step": 3647
    },
    {
      "epoch": 0.8684163542224603,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 3648
    },
    {
      "epoch": 0.8686544069511397,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 3649
    },
    {
      "epoch": 0.8688924596798191,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 3650
    },
    {
      "epoch": 0.8691305124084985,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2901,
      "step": 3651
    },
    {
      "epoch": 0.8693685651371779,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4451,
      "step": 3652
    },
    {
      "epoch": 0.8696066178658572,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 3653
    },
    {
      "epoch": 0.8698446705945367,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 3654
    },
    {
      "epoch": 0.8700827233232161,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 3655
    },
    {
      "epoch": 0.8703207760518955,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 3656
    },
    {
      "epoch": 0.8705588287805749,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5012,
      "step": 3657
    },
    {
      "epoch": 0.8707968815092543,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 3658
    },
    {
      "epoch": 0.8710349342379337,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4323,
      "step": 3659
    },
    {
      "epoch": 0.8712729869666131,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 3660
    },
    {
      "epoch": 0.8715110396952925,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 3661
    },
    {
      "epoch": 0.8717490924239719,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 3662
    },
    {
      "epoch": 0.8719871451526513,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 3663
    },
    {
      "epoch": 0.8722251978813307,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 3664
    },
    {
      "epoch": 0.8724632506100101,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2937,
      "step": 3665
    },
    {
      "epoch": 0.8727013033386896,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3672,
      "step": 3666
    },
    {
      "epoch": 0.8729393560673689,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 3667
    },
    {
      "epoch": 0.8731774087960483,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2698,
      "step": 3668
    },
    {
      "epoch": 0.8734154615247277,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 3669
    },
    {
      "epoch": 0.8736535142534071,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 3670
    },
    {
      "epoch": 0.8738915669820866,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 3671
    },
    {
      "epoch": 0.874129619710766,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 3672
    },
    {
      "epoch": 0.8743676724394454,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 3673
    },
    {
      "epoch": 0.8746057251681247,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3196,
      "step": 3674
    },
    {
      "epoch": 0.8748437778968041,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 3675
    },
    {
      "epoch": 0.8750818306254835,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 3676
    },
    {
      "epoch": 0.875319883354163,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 3677
    },
    {
      "epoch": 0.8755579360828424,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 3678
    },
    {
      "epoch": 0.8757959888115218,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3735,
      "step": 3679
    },
    {
      "epoch": 0.8760340415402011,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3165,
      "step": 3680
    },
    {
      "epoch": 0.8762720942688805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 3681
    },
    {
      "epoch": 0.8765101469975599,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 3682
    },
    {
      "epoch": 0.8767481997262394,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 3683
    },
    {
      "epoch": 0.8769862524549188,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 3684
    },
    {
      "epoch": 0.8772243051835982,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 3685
    },
    {
      "epoch": 0.8774623579122776,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 3686
    },
    {
      "epoch": 0.8777004106409569,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 3687
    },
    {
      "epoch": 0.8779384633696363,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 3688
    },
    {
      "epoch": 0.8781765160983158,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 3689
    },
    {
      "epoch": 0.8784145688269952,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.4601,
      "step": 3690
    },
    {
      "epoch": 0.8786526215556746,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2932,
      "step": 3691
    },
    {
      "epoch": 0.878890674284354,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 3692
    },
    {
      "epoch": 0.8791287270130334,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.4827,
      "step": 3693
    },
    {
      "epoch": 0.8793667797417127,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.5574,
      "step": 3694
    },
    {
      "epoch": 0.8796048324703922,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 3695
    },
    {
      "epoch": 0.8798428851990716,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 3696
    },
    {
      "epoch": 0.880080937927751,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 3697
    },
    {
      "epoch": 0.8803189906564304,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 3698
    },
    {
      "epoch": 0.8805570433851098,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 3699
    },
    {
      "epoch": 0.8807950961137893,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9133,
      "step": 3700
    },
    {
      "epoch": 0.8807950961137893,
      "eval_runtime": 0.5159,
      "eval_samples_per_second": 1.938,
      "eval_steps_per_second": 1.938,
      "step": 3700
    },
    {
      "epoch": 0.8810331488424686,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 3701
    },
    {
      "epoch": 0.881271201571148,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 3702
    },
    {
      "epoch": 0.8815092542998274,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 3703
    },
    {
      "epoch": 0.8817473070285068,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3908,
      "step": 3704
    },
    {
      "epoch": 0.8819853597571862,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 3705
    },
    {
      "epoch": 0.8822234124858657,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 3706
    },
    {
      "epoch": 0.882461465214545,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5075,
      "step": 3707
    },
    {
      "epoch": 0.8826995179432244,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 3708
    },
    {
      "epoch": 0.8829375706719038,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 3709
    },
    {
      "epoch": 0.8831756234005832,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 3710
    },
    {
      "epoch": 0.8834136761292626,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 3711
    },
    {
      "epoch": 0.8836517288579421,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 3712
    },
    {
      "epoch": 0.8838897815866215,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4281,
      "step": 3713
    },
    {
      "epoch": 0.8841278343153008,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 3714
    },
    {
      "epoch": 0.8843658870439802,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 3715
    },
    {
      "epoch": 0.8846039397726596,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 3716
    },
    {
      "epoch": 0.884841992501339,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 3717
    },
    {
      "epoch": 0.8850800452300185,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 3718
    },
    {
      "epoch": 0.8853180979586979,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 3719
    },
    {
      "epoch": 0.8855561506873773,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3222,
      "step": 3720
    },
    {
      "epoch": 0.8857942034160566,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 3721
    },
    {
      "epoch": 0.886032256144736,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 3722
    },
    {
      "epoch": 0.8862703088734155,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 3723
    },
    {
      "epoch": 0.8865083616020949,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4949,
      "step": 3724
    },
    {
      "epoch": 0.8867464143307743,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3507,
      "step": 3725
    },
    {
      "epoch": 0.8869844670594537,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 3726
    },
    {
      "epoch": 0.887222519788133,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 3727
    },
    {
      "epoch": 0.8874605725168124,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 3728
    },
    {
      "epoch": 0.8876986252454919,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 3729
    },
    {
      "epoch": 0.8879366779741713,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5283,
      "step": 3730
    },
    {
      "epoch": 0.8881747307028507,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 3731
    },
    {
      "epoch": 0.8884127834315301,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 3732
    },
    {
      "epoch": 0.8886508361602095,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 3733
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 3734
    },
    {
      "epoch": 0.8891269416175683,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 3735
    },
    {
      "epoch": 0.8893649943462477,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 3736
    },
    {
      "epoch": 0.8896030470749271,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 3737
    },
    {
      "epoch": 0.8898410998036065,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 3738
    },
    {
      "epoch": 0.8900791525322859,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 3739
    },
    {
      "epoch": 0.8903172052609652,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 3740
    },
    {
      "epoch": 0.8905552579896447,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 3741
    },
    {
      "epoch": 0.8907933107183241,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 3742
    },
    {
      "epoch": 0.8910313634470035,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 3743
    },
    {
      "epoch": 0.8912694161756829,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.6366,
      "step": 3744
    },
    {
      "epoch": 0.8915074689043623,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.3404,
      "step": 3745
    },
    {
      "epoch": 0.8917455216330418,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 3746
    },
    {
      "epoch": 0.8919835743617212,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 3747
    },
    {
      "epoch": 0.8922216270904005,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 3748
    },
    {
      "epoch": 0.8924596798190799,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3146,
      "step": 3749
    },
    {
      "epoch": 0.8926977325477593,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 3750
    },
    {
      "epoch": 0.8929357852764387,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4574,
      "step": 3751
    },
    {
      "epoch": 0.8931738380051182,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 3752
    },
    {
      "epoch": 0.8934118907337976,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 3753
    },
    {
      "epoch": 0.893649943462477,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 3754
    },
    {
      "epoch": 0.8938879961911563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 3755
    },
    {
      "epoch": 0.8941260489198357,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3573,
      "step": 3756
    },
    {
      "epoch": 0.8943641016485151,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4182,
      "step": 3757
    },
    {
      "epoch": 0.8946021543771946,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3442,
      "step": 3758
    },
    {
      "epoch": 0.894840207105874,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 3759
    },
    {
      "epoch": 0.8950782598345534,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 3760
    },
    {
      "epoch": 0.8953163125632327,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 3761
    },
    {
      "epoch": 0.8955543652919121,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 3762
    },
    {
      "epoch": 0.8957924180205915,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 3763
    },
    {
      "epoch": 0.896030470749271,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3878,
      "step": 3764
    },
    {
      "epoch": 0.8962685234779504,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 3765
    },
    {
      "epoch": 0.8965065762066298,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 3766
    },
    {
      "epoch": 0.8967446289353092,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3657,
      "step": 3767
    },
    {
      "epoch": 0.8969826816639885,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 3768
    },
    {
      "epoch": 0.897220734392668,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 3769
    },
    {
      "epoch": 0.8974587871213474,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 3770
    },
    {
      "epoch": 0.8976968398500268,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3409,
      "step": 3771
    },
    {
      "epoch": 0.8979348925787062,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 3772
    },
    {
      "epoch": 0.8981729453073856,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 3773
    },
    {
      "epoch": 0.898410998036065,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 3774
    },
    {
      "epoch": 0.8986490507647444,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 3775
    },
    {
      "epoch": 0.8988871034934238,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 3776
    },
    {
      "epoch": 0.8991251562221032,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 3777
    },
    {
      "epoch": 0.8993632089507826,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 3778
    },
    {
      "epoch": 0.899601261679462,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 3779
    },
    {
      "epoch": 0.8998393144081414,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 3780
    },
    {
      "epoch": 0.9000773671368208,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4704,
      "step": 3781
    },
    {
      "epoch": 0.9003154198655002,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 3782
    },
    {
      "epoch": 0.9005534725941796,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 3783
    },
    {
      "epoch": 0.900791525322859,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 3784
    },
    {
      "epoch": 0.9010295780515384,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 3785
    },
    {
      "epoch": 0.9012676307802178,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 3786
    },
    {
      "epoch": 0.9015056835088973,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 3787
    },
    {
      "epoch": 0.9017437362375766,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 3788
    },
    {
      "epoch": 0.901981788966256,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 3789
    },
    {
      "epoch": 0.9022198416949354,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 3790
    },
    {
      "epoch": 0.9024578944236148,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.238,
      "step": 3791
    },
    {
      "epoch": 0.9026959471522943,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 3792
    },
    {
      "epoch": 0.9029339998809737,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 3793
    },
    {
      "epoch": 0.903172052609653,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 3794
    },
    {
      "epoch": 0.9034101053383324,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 3795
    },
    {
      "epoch": 0.9036481580670118,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 3796
    },
    {
      "epoch": 0.9038862107956912,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4606,
      "step": 3797
    },
    {
      "epoch": 0.9041242635243707,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 3798
    },
    {
      "epoch": 0.9043623162530501,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 3799
    },
    {
      "epoch": 0.9046003689817295,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 3800
    },
    {
      "epoch": 0.9046003689817295,
      "eval_runtime": 0.5106,
      "eval_samples_per_second": 1.958,
      "eval_steps_per_second": 1.958,
      "step": 3800
    },
    {
      "epoch": 0.9048384217104088,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 3801
    },
    {
      "epoch": 0.9050764744390882,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 3802
    },
    {
      "epoch": 0.9053145271677676,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.548,
      "step": 3803
    },
    {
      "epoch": 0.9055525798964471,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3409,
      "step": 3804
    },
    {
      "epoch": 0.9057906326251265,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 3805
    },
    {
      "epoch": 0.9060286853538059,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3696,
      "step": 3806
    },
    {
      "epoch": 0.9062667380824853,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.5544,
      "step": 3807
    },
    {
      "epoch": 0.9065047908111646,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 3808
    },
    {
      "epoch": 0.906742843539844,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 3809
    },
    {
      "epoch": 0.9069808962685235,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 3810
    },
    {
      "epoch": 0.9072189489972029,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4416,
      "step": 3811
    },
    {
      "epoch": 0.9074570017258823,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4203,
      "step": 3812
    },
    {
      "epoch": 0.9076950544545617,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 3813
    },
    {
      "epoch": 0.907933107183241,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 3814
    },
    {
      "epoch": 0.9081711599119205,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 3815
    },
    {
      "epoch": 0.9084092126405999,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 3816
    },
    {
      "epoch": 0.9086472653692793,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 3817
    },
    {
      "epoch": 0.9088853180979587,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 3818
    },
    {
      "epoch": 0.9091233708266381,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 3819
    },
    {
      "epoch": 0.9093614235553175,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 3820
    },
    {
      "epoch": 0.909599476283997,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3917,
      "step": 3821
    },
    {
      "epoch": 0.9098375290126763,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8188,
      "step": 3822
    },
    {
      "epoch": 0.9100755817413557,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 3823
    },
    {
      "epoch": 0.9103136344700351,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4321,
      "step": 3824
    },
    {
      "epoch": 0.9105516871987145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3554,
      "step": 3825
    },
    {
      "epoch": 0.9107897399273939,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 3826
    },
    {
      "epoch": 0.9110277926560734,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4106,
      "step": 3827
    },
    {
      "epoch": 0.9112658453847527,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4425,
      "step": 3828
    },
    {
      "epoch": 0.9115038981134321,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.7058,
      "step": 3829
    },
    {
      "epoch": 0.9117419508421115,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 3830
    },
    {
      "epoch": 0.9119800035707909,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 3831
    },
    {
      "epoch": 0.9122180562994703,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 3832
    },
    {
      "epoch": 0.9124561090281498,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 3833
    },
    {
      "epoch": 0.9126941617568292,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 3834
    },
    {
      "epoch": 0.9129322144855085,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3717,
      "step": 3835
    },
    {
      "epoch": 0.9131702672141879,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 3836
    },
    {
      "epoch": 0.9134083199428673,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3065,
      "step": 3837
    },
    {
      "epoch": 0.9136463726715468,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 3838
    },
    {
      "epoch": 0.9138844254002262,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 3839
    },
    {
      "epoch": 0.9141224781289056,
      "grad_norm": 8.9375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 3840
    },
    {
      "epoch": 0.914360530857585,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 3841
    },
    {
      "epoch": 0.9145985835862643,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.5369,
      "step": 3842
    },
    {
      "epoch": 0.9148366363149437,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2283,
      "step": 3843
    },
    {
      "epoch": 0.9150746890436232,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 3844
    },
    {
      "epoch": 0.9153127417723026,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 3845
    },
    {
      "epoch": 0.915550794500982,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 3846
    },
    {
      "epoch": 0.9157888472296614,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 3847
    },
    {
      "epoch": 0.9160268999583407,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 3848
    },
    {
      "epoch": 0.9162649526870201,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3019,
      "step": 3849
    },
    {
      "epoch": 0.9165030054156996,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 3850
    },
    {
      "epoch": 0.916741058144379,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 3851
    },
    {
      "epoch": 0.9169791108730584,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4168,
      "step": 3852
    },
    {
      "epoch": 0.9172171636017378,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3536,
      "step": 3853
    },
    {
      "epoch": 0.9174552163304172,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 3854
    },
    {
      "epoch": 0.9176932690590965,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 3855
    },
    {
      "epoch": 0.917931321787776,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 3856
    },
    {
      "epoch": 0.9181693745164554,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 3857
    },
    {
      "epoch": 0.9184074272451348,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 3858
    },
    {
      "epoch": 0.9186454799738142,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 3859
    },
    {
      "epoch": 0.9188835327024936,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 3860
    },
    {
      "epoch": 0.919121585431173,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 3861
    },
    {
      "epoch": 0.9193596381598524,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 3862
    },
    {
      "epoch": 0.9195976908885318,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 3863
    },
    {
      "epoch": 0.9198357436172112,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 3864
    },
    {
      "epoch": 0.9200737963458906,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 3865
    },
    {
      "epoch": 0.92031184907457,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3664,
      "step": 3866
    },
    {
      "epoch": 0.9205499018032495,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 3867
    },
    {
      "epoch": 0.9207879545319289,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 3868
    },
    {
      "epoch": 0.9210260072606082,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 3869
    },
    {
      "epoch": 0.9212640599892876,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 3870
    },
    {
      "epoch": 0.921502112717967,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5068,
      "step": 3871
    },
    {
      "epoch": 0.9217401654466464,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 3872
    },
    {
      "epoch": 0.9219782181753259,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 3873
    },
    {
      "epoch": 0.9222162709040053,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.4084,
      "step": 3874
    },
    {
      "epoch": 0.9224543236326846,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.327,
      "step": 3875
    },
    {
      "epoch": 0.922692376361364,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 3876
    },
    {
      "epoch": 0.9229304290900434,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 3877
    },
    {
      "epoch": 0.9231684818187228,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4887,
      "step": 3878
    },
    {
      "epoch": 0.9234065345474023,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 3879
    },
    {
      "epoch": 0.9236445872760817,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 3880
    },
    {
      "epoch": 0.9238826400047611,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 3881
    },
    {
      "epoch": 0.9241206927334404,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 3882
    },
    {
      "epoch": 0.9243587454621198,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 3883
    },
    {
      "epoch": 0.9245967981907992,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3575,
      "step": 3884
    },
    {
      "epoch": 0.9248348509194787,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 3885
    },
    {
      "epoch": 0.9250729036481581,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2474,
      "step": 3886
    },
    {
      "epoch": 0.9253109563768375,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3062,
      "step": 3887
    },
    {
      "epoch": 0.9255490091055169,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 3888
    },
    {
      "epoch": 0.9257870618341962,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3496,
      "step": 3889
    },
    {
      "epoch": 0.9260251145628757,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 3890
    },
    {
      "epoch": 0.9262631672915551,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 3891
    },
    {
      "epoch": 0.9265012200202345,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 3892
    },
    {
      "epoch": 0.9267392727489139,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 3893
    },
    {
      "epoch": 0.9269773254775933,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 3894
    },
    {
      "epoch": 0.9272153782062726,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 3895
    },
    {
      "epoch": 0.9274534309349521,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3921,
      "step": 3896
    },
    {
      "epoch": 0.9276914836636315,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 3897
    },
    {
      "epoch": 0.9279295363923109,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4341,
      "step": 3898
    },
    {
      "epoch": 0.9281675891209903,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 3899
    },
    {
      "epoch": 0.9284056418496697,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.46,
      "step": 3900
    },
    {
      "epoch": 0.9284056418496697,
      "eval_runtime": 0.5019,
      "eval_samples_per_second": 1.993,
      "eval_steps_per_second": 1.993,
      "step": 3900
    },
    {
      "epoch": 0.928643694578349,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 3901
    },
    {
      "epoch": 0.9288817473070285,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 3902
    },
    {
      "epoch": 0.9291198000357079,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 3903
    },
    {
      "epoch": 0.9293578527643873,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 3904
    },
    {
      "epoch": 0.9295959054930667,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 3905
    },
    {
      "epoch": 0.9298339582217461,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 3906
    },
    {
      "epoch": 0.9300720109504255,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 3907
    },
    {
      "epoch": 0.930310063679105,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.5618,
      "step": 3908
    },
    {
      "epoch": 0.9305481164077843,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 3909
    },
    {
      "epoch": 0.9307861691364637,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 3910
    },
    {
      "epoch": 0.9310242218651431,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 3911
    },
    {
      "epoch": 0.9312622745938225,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3801,
      "step": 3912
    },
    {
      "epoch": 0.931500327322502,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 3913
    },
    {
      "epoch": 0.9317383800511814,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 3914
    },
    {
      "epoch": 0.9319764327798608,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 3915
    },
    {
      "epoch": 0.9322144855085401,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3139,
      "step": 3916
    },
    {
      "epoch": 0.9324525382372195,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 3917
    },
    {
      "epoch": 0.9326905909658989,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3605,
      "step": 3918
    },
    {
      "epoch": 0.9329286436945784,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 3919
    },
    {
      "epoch": 0.9331666964232578,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 3920
    },
    {
      "epoch": 0.9334047491519372,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.433,
      "step": 3921
    },
    {
      "epoch": 0.9336428018806165,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 3922
    },
    {
      "epoch": 0.9338808546092959,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4725,
      "step": 3923
    },
    {
      "epoch": 0.9341189073379753,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 3924
    },
    {
      "epoch": 0.9343569600666548,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 3925
    },
    {
      "epoch": 0.9345950127953342,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 3926
    },
    {
      "epoch": 0.9348330655240136,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 3927
    },
    {
      "epoch": 0.935071118252693,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 3928
    },
    {
      "epoch": 0.9353091709813723,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3178,
      "step": 3929
    },
    {
      "epoch": 0.9355472237100517,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 3930
    },
    {
      "epoch": 0.9357852764387312,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 3931
    },
    {
      "epoch": 0.9360233291674106,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 3932
    },
    {
      "epoch": 0.93626138189609,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 3933
    },
    {
      "epoch": 0.9364994346247694,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 3934
    },
    {
      "epoch": 0.9367374873534488,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3335,
      "step": 3935
    },
    {
      "epoch": 0.9369755400821282,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 3936
    },
    {
      "epoch": 0.9372135928108076,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 3937
    },
    {
      "epoch": 0.937451645539487,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 3938
    },
    {
      "epoch": 0.9376896982681664,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 3939
    },
    {
      "epoch": 0.9379277509968458,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 3940
    },
    {
      "epoch": 0.9381658037255252,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 3941
    },
    {
      "epoch": 0.9384038564542047,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 3942
    },
    {
      "epoch": 0.938641909182884,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 3943
    },
    {
      "epoch": 0.9388799619115634,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 3944
    },
    {
      "epoch": 0.9391180146402428,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 3945
    },
    {
      "epoch": 0.9393560673689222,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 3946
    },
    {
      "epoch": 0.9395941200976016,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3264,
      "step": 3947
    },
    {
      "epoch": 0.9398321728262811,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4389,
      "step": 3948
    },
    {
      "epoch": 0.9400702255549604,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 3949
    },
    {
      "epoch": 0.9403082782836398,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 3950
    },
    {
      "epoch": 0.9405463310123192,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 3951
    },
    {
      "epoch": 0.9407843837409986,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 3952
    },
    {
      "epoch": 0.941022436469678,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 3953
    },
    {
      "epoch": 0.9412604891983575,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 3954
    },
    {
      "epoch": 0.9414985419270369,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 3955
    },
    {
      "epoch": 0.9417365946557162,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 3956
    },
    {
      "epoch": 0.9419746473843956,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4877,
      "step": 3957
    },
    {
      "epoch": 0.942212700113075,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.6115,
      "step": 3958
    },
    {
      "epoch": 0.9424507528417545,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 3959
    },
    {
      "epoch": 0.9426888055704339,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 3960
    },
    {
      "epoch": 0.9429268582991133,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 3961
    },
    {
      "epoch": 0.9431649110277927,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 3962
    },
    {
      "epoch": 0.943402963756472,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5791,
      "step": 3963
    },
    {
      "epoch": 0.9436410164851514,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 3964
    },
    {
      "epoch": 0.9438790692138309,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 3965
    },
    {
      "epoch": 0.9441171219425103,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 3966
    },
    {
      "epoch": 0.9443551746711897,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 3967
    },
    {
      "epoch": 0.9445932273998691,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3502,
      "step": 3968
    },
    {
      "epoch": 0.9448312801285484,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 3969
    },
    {
      "epoch": 0.9450693328572278,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 3970
    },
    {
      "epoch": 0.9453073855859073,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 3971
    },
    {
      "epoch": 0.9455454383145867,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 3972
    },
    {
      "epoch": 0.9457834910432661,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3302,
      "step": 3973
    },
    {
      "epoch": 0.9460215437719455,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 3974
    },
    {
      "epoch": 0.9462595965006249,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 3975
    },
    {
      "epoch": 0.9464976492293042,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2883,
      "step": 3976
    },
    {
      "epoch": 0.9467357019579837,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3275,
      "step": 3977
    },
    {
      "epoch": 0.9469737546866631,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 3978
    },
    {
      "epoch": 0.9472118074153425,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3748,
      "step": 3979
    },
    {
      "epoch": 0.9474498601440219,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 3980
    },
    {
      "epoch": 0.9476879128727013,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 3981
    },
    {
      "epoch": 0.9479259656013808,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 3982
    },
    {
      "epoch": 0.9481640183300601,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 3983
    },
    {
      "epoch": 0.9484020710587395,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 3984
    },
    {
      "epoch": 0.9486401237874189,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 3985
    },
    {
      "epoch": 0.9488781765160983,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 3986
    },
    {
      "epoch": 0.9491162292447777,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 3987
    },
    {
      "epoch": 0.9493542819734572,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 3988
    },
    {
      "epoch": 0.9495923347021366,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3397,
      "step": 3989
    },
    {
      "epoch": 0.9498303874308159,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 3990
    },
    {
      "epoch": 0.9500684401594953,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3843,
      "step": 3991
    },
    {
      "epoch": 0.9503064928881747,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 3992
    },
    {
      "epoch": 0.9505445456168541,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 3993
    },
    {
      "epoch": 0.9507825983455336,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 3994
    },
    {
      "epoch": 0.951020651074213,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1747,
      "step": 3995
    },
    {
      "epoch": 0.9512587038028923,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 3996
    },
    {
      "epoch": 0.9514967565315717,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4714,
      "step": 3997
    },
    {
      "epoch": 0.9517348092602511,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5588,
      "step": 3998
    },
    {
      "epoch": 0.9519728619889305,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 3999
    },
    {
      "epoch": 0.95221091471761,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 4000
    },
    {
      "epoch": 0.95221091471761,
      "eval_runtime": 0.5037,
      "eval_samples_per_second": 1.985,
      "eval_steps_per_second": 1.985,
      "step": 4000
    },
    {
      "epoch": 0.9524489674462894,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3697,
      "step": 4001
    },
    {
      "epoch": 0.9526870201749688,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2689,
      "step": 4002
    },
    {
      "epoch": 0.9529250729036481,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3805,
      "step": 4003
    },
    {
      "epoch": 0.9531631256323275,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 4004
    },
    {
      "epoch": 0.953401178361007,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 4005
    },
    {
      "epoch": 0.9536392310896864,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 4006
    },
    {
      "epoch": 0.9538772838183658,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 4007
    },
    {
      "epoch": 0.9541153365470452,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 4008
    },
    {
      "epoch": 0.9543533892757246,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 4009
    },
    {
      "epoch": 0.9545914420044039,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 4010
    },
    {
      "epoch": 0.9548294947330834,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 4011
    },
    {
      "epoch": 0.9550675474617628,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 4012
    },
    {
      "epoch": 0.9553056001904422,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 4013
    },
    {
      "epoch": 0.9555436529191216,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3721,
      "step": 4014
    },
    {
      "epoch": 0.955781705647801,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 4015
    },
    {
      "epoch": 0.9560197583764803,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4561,
      "step": 4016
    },
    {
      "epoch": 0.9562578111051598,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5725,
      "step": 4017
    },
    {
      "epoch": 0.9564958638338392,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2572,
      "step": 4018
    },
    {
      "epoch": 0.9567339165625186,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 4019
    },
    {
      "epoch": 0.956971969291198,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 4020
    },
    {
      "epoch": 0.9572100220198774,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 4021
    },
    {
      "epoch": 0.9574480747485568,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4357,
      "step": 4022
    },
    {
      "epoch": 0.9576861274772362,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 4023
    },
    {
      "epoch": 0.9579241802059156,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4763,
      "step": 4024
    },
    {
      "epoch": 0.958162232934595,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 4025
    },
    {
      "epoch": 0.9584002856632744,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 4026
    },
    {
      "epoch": 0.9586383383919538,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 4027
    },
    {
      "epoch": 0.9588763911206333,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 4028
    },
    {
      "epoch": 0.9591144438493127,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3935,
      "step": 4029
    },
    {
      "epoch": 0.959352496577992,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 4030
    },
    {
      "epoch": 0.9595905493066714,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3953,
      "step": 4031
    },
    {
      "epoch": 0.9598286020353508,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4324,
      "step": 4032
    },
    {
      "epoch": 0.9600666547640302,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 4033
    },
    {
      "epoch": 0.9603047074927097,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 4034
    },
    {
      "epoch": 0.9605427602213891,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 4035
    },
    {
      "epoch": 0.9607808129500685,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 4036
    },
    {
      "epoch": 0.9610188656787478,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 4037
    },
    {
      "epoch": 0.9612569184074272,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 4038
    },
    {
      "epoch": 0.9614949711361066,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4014,
      "step": 4039
    },
    {
      "epoch": 0.9617330238647861,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 4040
    },
    {
      "epoch": 0.9619710765934655,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 4041
    },
    {
      "epoch": 0.9622091293221449,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5201,
      "step": 4042
    },
    {
      "epoch": 0.9624471820508242,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2674,
      "step": 4043
    },
    {
      "epoch": 0.9626852347795036,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 4044
    },
    {
      "epoch": 0.962923287508183,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2611,
      "step": 4045
    },
    {
      "epoch": 0.9631613402368625,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 4046
    },
    {
      "epoch": 0.9633993929655419,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 4047
    },
    {
      "epoch": 0.9636374456942213,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4982,
      "step": 4048
    },
    {
      "epoch": 0.9638754984229007,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3983,
      "step": 4049
    },
    {
      "epoch": 0.96411355115158,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2923,
      "step": 4050
    },
    {
      "epoch": 0.9643516038802594,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 4051
    },
    {
      "epoch": 0.9645896566089389,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 4052
    },
    {
      "epoch": 0.9648277093376183,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 4053
    },
    {
      "epoch": 0.9650657620662977,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 4054
    },
    {
      "epoch": 0.9653038147949771,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 4055
    },
    {
      "epoch": 0.9655418675236564,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 4056
    },
    {
      "epoch": 0.9657799202523359,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 4057
    },
    {
      "epoch": 0.9660179729810153,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 4058
    },
    {
      "epoch": 0.9662560257096947,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.4162,
      "step": 4059
    },
    {
      "epoch": 0.9664940784383741,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.413,
      "step": 4060
    },
    {
      "epoch": 0.9667321311670535,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 4061
    },
    {
      "epoch": 0.9669701838957329,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 4062
    },
    {
      "epoch": 0.9672082366244124,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 4063
    },
    {
      "epoch": 0.9674462893530917,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 4064
    },
    {
      "epoch": 0.9676843420817711,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 4065
    },
    {
      "epoch": 0.9679223948104505,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 4066
    },
    {
      "epoch": 0.9681604475391299,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 4067
    },
    {
      "epoch": 0.9683985002678093,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 4068
    },
    {
      "epoch": 0.9686365529964888,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3551,
      "step": 4069
    },
    {
      "epoch": 0.9688746057251681,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 4070
    },
    {
      "epoch": 0.9691126584538475,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 4071
    },
    {
      "epoch": 0.9693507111825269,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 4072
    },
    {
      "epoch": 0.9695887639112063,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 4073
    },
    {
      "epoch": 0.9698268166398857,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 4074
    },
    {
      "epoch": 0.9700648693685652,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 4075
    },
    {
      "epoch": 0.9703029220972446,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 4076
    },
    {
      "epoch": 0.9705409748259239,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 4077
    },
    {
      "epoch": 0.9707790275546033,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 4078
    },
    {
      "epoch": 0.9710170802832827,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 4079
    },
    {
      "epoch": 0.9712551330119622,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.4444,
      "step": 4080
    },
    {
      "epoch": 0.9714931857406416,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3609,
      "step": 4081
    },
    {
      "epoch": 0.971731238469321,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 4082
    },
    {
      "epoch": 0.9719692911980004,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3829,
      "step": 4083
    },
    {
      "epoch": 0.9722073439266797,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 4084
    },
    {
      "epoch": 0.9724453966553591,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 4085
    },
    {
      "epoch": 0.9726834493840386,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 4086
    },
    {
      "epoch": 0.972921502112718,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3467,
      "step": 4087
    },
    {
      "epoch": 0.9731595548413974,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 4088
    },
    {
      "epoch": 0.9733976075700768,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 4089
    },
    {
      "epoch": 0.9736356602987561,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 4090
    },
    {
      "epoch": 0.9738737130274355,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 4091
    },
    {
      "epoch": 0.974111765756115,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3318,
      "step": 4092
    },
    {
      "epoch": 0.9743498184847944,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5352,
      "step": 4093
    },
    {
      "epoch": 0.9745878712134738,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 4094
    },
    {
      "epoch": 0.9748259239421532,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 4095
    },
    {
      "epoch": 0.9750639766708326,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.4897,
      "step": 4096
    },
    {
      "epoch": 0.9753020293995119,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4086,
      "step": 4097
    },
    {
      "epoch": 0.9755400821281914,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 4098
    },
    {
      "epoch": 0.9757781348568708,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 4099
    },
    {
      "epoch": 0.9760161875855502,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 4100
    },
    {
      "epoch": 0.9760161875855502,
      "eval_runtime": 0.5178,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 4100
    },
    {
      "epoch": 0.9762542403142296,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 4101
    },
    {
      "epoch": 0.976492293042909,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4352,
      "step": 4102
    },
    {
      "epoch": 0.9767303457715885,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 4103
    },
    {
      "epoch": 0.9769683985002678,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 4104
    },
    {
      "epoch": 0.9772064512289472,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 4105
    },
    {
      "epoch": 0.9774445039576266,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 4106
    },
    {
      "epoch": 0.977682556686306,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2943,
      "step": 4107
    },
    {
      "epoch": 0.9779206094149854,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4659,
      "step": 4108
    },
    {
      "epoch": 0.9781586621436649,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 4109
    },
    {
      "epoch": 0.9783967148723443,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 4110
    },
    {
      "epoch": 0.9786347676010236,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 4111
    },
    {
      "epoch": 0.978872820329703,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 4112
    },
    {
      "epoch": 0.9791108730583824,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2894,
      "step": 4113
    },
    {
      "epoch": 0.9793489257870618,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 4114
    },
    {
      "epoch": 0.9795869785157413,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 4115
    },
    {
      "epoch": 0.9798250312444207,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 4116
    },
    {
      "epoch": 0.9800630839731,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 4117
    },
    {
      "epoch": 0.9803011367017794,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 4118
    },
    {
      "epoch": 0.9805391894304588,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 4119
    },
    {
      "epoch": 0.9807772421591382,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 4120
    },
    {
      "epoch": 0.9810152948878177,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4375,
      "step": 4121
    },
    {
      "epoch": 0.9812533476164971,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 4122
    },
    {
      "epoch": 0.9814914003451765,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 4123
    },
    {
      "epoch": 0.9817294530738558,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2804,
      "step": 4124
    },
    {
      "epoch": 0.9819675058025352,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 4125
    },
    {
      "epoch": 0.9822055585312147,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 4126
    },
    {
      "epoch": 0.9824436112598941,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 4127
    },
    {
      "epoch": 0.9826816639885735,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2258,
      "step": 4128
    },
    {
      "epoch": 0.9829197167172529,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3714,
      "step": 4129
    },
    {
      "epoch": 0.9831577694459323,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 4130
    },
    {
      "epoch": 0.9833958221746116,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 4131
    },
    {
      "epoch": 0.9836338749032911,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 4132
    },
    {
      "epoch": 0.9838719276319705,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3685,
      "step": 4133
    },
    {
      "epoch": 0.9841099803606499,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 4134
    },
    {
      "epoch": 0.9843480330893293,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 4135
    },
    {
      "epoch": 0.9845860858180087,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 4136
    },
    {
      "epoch": 0.984824138546688,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 4137
    },
    {
      "epoch": 0.9850621912753675,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 4138
    },
    {
      "epoch": 0.9853002440040469,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3754,
      "step": 4139
    },
    {
      "epoch": 0.9855382967327263,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 4140
    },
    {
      "epoch": 0.9857763494614057,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 4141
    },
    {
      "epoch": 0.9860144021900851,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 4142
    },
    {
      "epoch": 0.9862524549187645,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3485,
      "step": 4143
    },
    {
      "epoch": 0.986490507647444,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 4144
    },
    {
      "epoch": 0.9867285603761233,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 4145
    },
    {
      "epoch": 0.9869666131048027,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 4146
    },
    {
      "epoch": 0.9872046658334821,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4013,
      "step": 4147
    },
    {
      "epoch": 0.9874427185621615,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4294,
      "step": 4148
    },
    {
      "epoch": 0.987680771290841,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 4149
    },
    {
      "epoch": 0.9879188240195204,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 4150
    },
    {
      "epoch": 0.9881568767481997,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4592,
      "step": 4151
    },
    {
      "epoch": 0.9883949294768791,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 4152
    },
    {
      "epoch": 0.9886329822055585,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 4153
    },
    {
      "epoch": 0.9888710349342379,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4097,
      "step": 4154
    },
    {
      "epoch": 0.9891090876629174,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 4155
    },
    {
      "epoch": 0.9893471403915968,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 4156
    },
    {
      "epoch": 0.9895851931202762,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4508,
      "step": 4157
    },
    {
      "epoch": 0.9898232458489555,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 4158
    },
    {
      "epoch": 0.9900612985776349,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4607,
      "step": 4159
    },
    {
      "epoch": 0.9902993513063143,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 4160
    },
    {
      "epoch": 0.9905374040349938,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.7879,
      "step": 4161
    },
    {
      "epoch": 0.9907754567636732,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 4162
    },
    {
      "epoch": 0.9910135094923526,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 4163
    },
    {
      "epoch": 0.991251562221032,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.313,
      "step": 4164
    },
    {
      "epoch": 0.9914896149497113,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 4165
    },
    {
      "epoch": 0.9917276676783907,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4699,
      "step": 4166
    },
    {
      "epoch": 0.9919657204070702,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 4167
    },
    {
      "epoch": 0.9922037731357496,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 4168
    },
    {
      "epoch": 0.992441825864429,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5617,
      "step": 4169
    },
    {
      "epoch": 0.9926798785931084,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 4170
    },
    {
      "epoch": 0.9929179313217877,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 4171
    },
    {
      "epoch": 0.9931559840504672,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 4172
    },
    {
      "epoch": 0.9933940367791466,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2629,
      "step": 4173
    },
    {
      "epoch": 0.993632089507826,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 4174
    },
    {
      "epoch": 0.9938701422365054,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 4175
    },
    {
      "epoch": 0.9941081949651848,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 4176
    },
    {
      "epoch": 0.9943462476938641,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 4177
    },
    {
      "epoch": 0.9945843004225436,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 4178
    },
    {
      "epoch": 0.994822353151223,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 4179
    },
    {
      "epoch": 0.9950604058799024,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 4180
    },
    {
      "epoch": 0.9952984586085818,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 4181
    },
    {
      "epoch": 0.9955365113372612,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 4182
    },
    {
      "epoch": 0.9957745640659406,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 4183
    },
    {
      "epoch": 0.99601261679462,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 4184
    },
    {
      "epoch": 0.9962506695232994,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 4185
    },
    {
      "epoch": 0.9964887222519788,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 4186
    },
    {
      "epoch": 0.9967267749806582,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 4187
    },
    {
      "epoch": 0.9969648277093376,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.379,
      "step": 4188
    },
    {
      "epoch": 0.997202880438017,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 4189
    },
    {
      "epoch": 0.9974409331666965,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 4190
    },
    {
      "epoch": 0.9976789858953758,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 4191
    },
    {
      "epoch": 0.9979170386240552,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 4192
    },
    {
      "epoch": 0.9981550913527346,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 4193
    },
    {
      "epoch": 0.998393144081414,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 4194
    },
    {
      "epoch": 0.9986311968100935,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 4195
    },
    {
      "epoch": 0.9988692495387729,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3396,
      "step": 4196
    },
    {
      "epoch": 0.9991073022674523,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 4197
    },
    {
      "epoch": 0.9993453549961316,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 4198
    },
    {
      "epoch": 0.999583407724811,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 4199
    },
    {
      "epoch": 0.9998214604534904,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 4200
    },
    {
      "epoch": 0.9998214604534904,
      "eval_runtime": 0.5316,
      "eval_samples_per_second": 1.881,
      "eval_steps_per_second": 1.881,
      "step": 4200
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 4201
    },
    {
      "epoch": 1.0002380527286794,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 4202
    },
    {
      "epoch": 1.0004761054573588,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 4203
    },
    {
      "epoch": 1.0007141581860381,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 4204
    },
    {
      "epoch": 1.0009522109147175,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.394,
      "step": 4205
    },
    {
      "epoch": 1.001190263643397,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 4206
    },
    {
      "epoch": 1.0014283163720765,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 4207
    },
    {
      "epoch": 1.001666369100756,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 4208
    },
    {
      "epoch": 1.0019044218294353,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 4209
    },
    {
      "epoch": 1.0021424745581147,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.4399,
      "step": 4210
    },
    {
      "epoch": 1.002380527286794,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 4211
    },
    {
      "epoch": 1.0026185800154734,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 4212
    },
    {
      "epoch": 1.0028566327441528,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 4213
    },
    {
      "epoch": 1.0030946854728322,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 4214
    },
    {
      "epoch": 1.0033327382015116,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 4215
    },
    {
      "epoch": 1.003570790930191,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 4216
    },
    {
      "epoch": 1.0038088436588704,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5012,
      "step": 4217
    },
    {
      "epoch": 1.0040468963875497,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 4218
    },
    {
      "epoch": 1.0042849491162293,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 4219
    },
    {
      "epoch": 1.0045230018449087,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 4220
    },
    {
      "epoch": 1.004761054573588,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 4221
    },
    {
      "epoch": 1.0049991073022675,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 4222
    },
    {
      "epoch": 1.0052371600309469,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 4223
    },
    {
      "epoch": 1.0054752127596263,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 4224
    },
    {
      "epoch": 1.0057132654883056,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 4225
    },
    {
      "epoch": 1.005951318216985,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3566,
      "step": 4226
    },
    {
      "epoch": 1.0061893709456644,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 4227
    },
    {
      "epoch": 1.0064274236743438,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 4228
    },
    {
      "epoch": 1.0066654764030232,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 4229
    },
    {
      "epoch": 1.0069035291317028,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 4230
    },
    {
      "epoch": 1.0071415818603822,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5299,
      "step": 4231
    },
    {
      "epoch": 1.0073796345890615,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3828,
      "step": 4232
    },
    {
      "epoch": 1.007617687317741,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 4233
    },
    {
      "epoch": 1.0078557400464203,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 4234
    },
    {
      "epoch": 1.0080937927750997,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 4235
    },
    {
      "epoch": 1.008331845503779,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 4236
    },
    {
      "epoch": 1.0085698982324585,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 4237
    },
    {
      "epoch": 1.0088079509611378,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 4238
    },
    {
      "epoch": 1.0090460036898172,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7234,
      "step": 4239
    },
    {
      "epoch": 1.0092840564184966,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 4240
    },
    {
      "epoch": 1.009522109147176,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3933,
      "step": 4241
    },
    {
      "epoch": 1.0097601618758556,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4162,
      "step": 4242
    },
    {
      "epoch": 1.009998214604535,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 4243
    },
    {
      "epoch": 1.0102362673332144,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 4244
    },
    {
      "epoch": 1.0104743200618937,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 4245
    },
    {
      "epoch": 1.0107123727905731,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 4246
    },
    {
      "epoch": 1.0109504255192525,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 4247
    },
    {
      "epoch": 1.011188478247932,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 4248
    },
    {
      "epoch": 1.0114265309766113,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 4249
    },
    {
      "epoch": 1.0116645837052907,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5345,
      "step": 4250
    },
    {
      "epoch": 1.01190263643397,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 4251
    },
    {
      "epoch": 1.0121406891626494,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 4252
    },
    {
      "epoch": 1.012378741891329,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 4253
    },
    {
      "epoch": 1.0126167946200084,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 4254
    },
    {
      "epoch": 1.0128548473486878,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 4255
    },
    {
      "epoch": 1.0130929000773672,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 4256
    },
    {
      "epoch": 1.0133309528060466,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 4257
    },
    {
      "epoch": 1.013569005534726,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.459,
      "step": 4258
    },
    {
      "epoch": 1.0138070582634053,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2894,
      "step": 4259
    },
    {
      "epoch": 1.0140451109920847,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 4260
    },
    {
      "epoch": 1.014283163720764,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 4261
    },
    {
      "epoch": 1.0145212164494435,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 4262
    },
    {
      "epoch": 1.0147592691781229,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 4263
    },
    {
      "epoch": 1.0149973219068023,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 4264
    },
    {
      "epoch": 1.0152353746354819,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 4265
    },
    {
      "epoch": 1.0154734273641612,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4697,
      "step": 4266
    },
    {
      "epoch": 1.0157114800928406,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 4267
    },
    {
      "epoch": 1.01594953282152,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3124,
      "step": 4268
    },
    {
      "epoch": 1.0161875855501994,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 4269
    },
    {
      "epoch": 1.0164256382788788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3918,
      "step": 4270
    },
    {
      "epoch": 1.0166636910075582,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2312,
      "step": 4271
    },
    {
      "epoch": 1.0169017437362375,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 4272
    },
    {
      "epoch": 1.017139796464917,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 4273
    },
    {
      "epoch": 1.0173778491935963,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 4274
    },
    {
      "epoch": 1.0176159019222757,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 4275
    },
    {
      "epoch": 1.0178539546509553,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 4276
    },
    {
      "epoch": 1.0180920073796347,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 4277
    },
    {
      "epoch": 1.018330060108314,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 4278
    },
    {
      "epoch": 1.0185681128369934,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.6805,
      "step": 4279
    },
    {
      "epoch": 1.0188061655656728,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 4280
    },
    {
      "epoch": 1.0190442182943522,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 4281
    },
    {
      "epoch": 1.0192822710230316,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 4282
    },
    {
      "epoch": 1.019520323751711,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3308,
      "step": 4283
    },
    {
      "epoch": 1.0197583764803904,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 4284
    },
    {
      "epoch": 1.0199964292090697,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 4285
    },
    {
      "epoch": 1.0202344819377491,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.5224,
      "step": 4286
    },
    {
      "epoch": 1.0204725346664285,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 4287
    },
    {
      "epoch": 1.0207105873951081,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 4288
    },
    {
      "epoch": 1.0209486401237875,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 4289
    },
    {
      "epoch": 1.0211866928524669,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 4290
    },
    {
      "epoch": 1.0214247455811463,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 4291
    },
    {
      "epoch": 1.0216627983098256,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3837,
      "step": 4292
    },
    {
      "epoch": 1.021900851038505,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 4293
    },
    {
      "epoch": 1.0221389037671844,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.5264,
      "step": 4294
    },
    {
      "epoch": 1.0223769564958638,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 4295
    },
    {
      "epoch": 1.0226150092245432,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.4518,
      "step": 4296
    },
    {
      "epoch": 1.0228530619532226,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 4297
    },
    {
      "epoch": 1.023091114681902,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 4298
    },
    {
      "epoch": 1.0233291674105816,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 4299
    },
    {
      "epoch": 1.023567220139261,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 4300
    },
    {
      "epoch": 1.023567220139261,
      "eval_runtime": 0.5165,
      "eval_samples_per_second": 1.936,
      "eval_steps_per_second": 1.936,
      "step": 4300
    },
    {
      "epoch": 1.0238052728679403,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 4301
    },
    {
      "epoch": 1.0240433255966197,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 4302
    },
    {
      "epoch": 1.024281378325299,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 4303
    },
    {
      "epoch": 1.0245194310539785,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 4304
    },
    {
      "epoch": 1.0247574837826579,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3142,
      "step": 4305
    },
    {
      "epoch": 1.0249955365113372,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3639,
      "step": 4306
    },
    {
      "epoch": 1.0252335892400166,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3527,
      "step": 4307
    },
    {
      "epoch": 1.025471641968696,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 4308
    },
    {
      "epoch": 1.0257096946973754,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 4309
    },
    {
      "epoch": 1.0259477474260548,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3498,
      "step": 4310
    },
    {
      "epoch": 1.0261858001547344,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 4311
    },
    {
      "epoch": 1.0264238528834138,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4151,
      "step": 4312
    },
    {
      "epoch": 1.0266619056120931,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 4313
    },
    {
      "epoch": 1.0268999583407725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2471,
      "step": 4314
    },
    {
      "epoch": 1.027138011069452,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 4315
    },
    {
      "epoch": 1.0273760637981313,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 4316
    },
    {
      "epoch": 1.0276141165268107,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 4317
    },
    {
      "epoch": 1.02785216925549,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 4318
    },
    {
      "epoch": 1.0280902219841694,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 4319
    },
    {
      "epoch": 1.0283282747128488,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 4320
    },
    {
      "epoch": 1.0285663274415282,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.342,
      "step": 4321
    },
    {
      "epoch": 1.0288043801702078,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 4322
    },
    {
      "epoch": 1.0290424328988872,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4504,
      "step": 4323
    },
    {
      "epoch": 1.0292804856275666,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.272,
      "step": 4324
    },
    {
      "epoch": 1.029518538356246,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 4325
    },
    {
      "epoch": 1.0297565910849253,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 4326
    },
    {
      "epoch": 1.0299946438136047,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2749,
      "step": 4327
    },
    {
      "epoch": 1.030232696542284,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 4328
    },
    {
      "epoch": 1.0304707492709635,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 4329
    },
    {
      "epoch": 1.0307088019996429,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 4330
    },
    {
      "epoch": 1.0309468547283223,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 4331
    },
    {
      "epoch": 1.0311849074570016,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2528,
      "step": 4332
    },
    {
      "epoch": 1.031422960185681,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 4333
    },
    {
      "epoch": 1.0316610129143606,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 4334
    },
    {
      "epoch": 1.03189906564304,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 4335
    },
    {
      "epoch": 1.0321371183717194,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3076,
      "step": 4336
    },
    {
      "epoch": 1.0323751711003988,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 4337
    },
    {
      "epoch": 1.0326132238290782,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 4338
    },
    {
      "epoch": 1.0328512765577575,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 4339
    },
    {
      "epoch": 1.033089329286437,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 4340
    },
    {
      "epoch": 1.0333273820151163,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 4341
    },
    {
      "epoch": 1.0335654347437957,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 4342
    },
    {
      "epoch": 1.033803487472475,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3735,
      "step": 4343
    },
    {
      "epoch": 1.0340415402011545,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 4344
    },
    {
      "epoch": 1.0342795929298338,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 4345
    },
    {
      "epoch": 1.0345176456585135,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 4346
    },
    {
      "epoch": 1.0347556983871928,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 4347
    },
    {
      "epoch": 1.0349937511158722,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2834,
      "step": 4348
    },
    {
      "epoch": 1.0352318038445516,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 4349
    },
    {
      "epoch": 1.035469856573231,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 4350
    },
    {
      "epoch": 1.0357079093019104,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 4351
    },
    {
      "epoch": 1.0359459620305898,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 4352
    },
    {
      "epoch": 1.0361840147592691,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 4353
    },
    {
      "epoch": 1.0364220674879485,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 4354
    },
    {
      "epoch": 1.036660120216628,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 4355
    },
    {
      "epoch": 1.0368981729453073,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 4356
    },
    {
      "epoch": 1.037136225673987,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2694,
      "step": 4357
    },
    {
      "epoch": 1.0373742784026663,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 4358
    },
    {
      "epoch": 1.0376123311313457,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 4359
    },
    {
      "epoch": 1.037850383860025,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 4360
    },
    {
      "epoch": 1.0380884365887044,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 4361
    },
    {
      "epoch": 1.0383264893173838,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 4362
    },
    {
      "epoch": 1.0385645420460632,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 4363
    },
    {
      "epoch": 1.0388025947747426,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 4364
    },
    {
      "epoch": 1.039040647503422,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 4365
    },
    {
      "epoch": 1.0392787002321013,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 4366
    },
    {
      "epoch": 1.0395167529607807,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.5335,
      "step": 4367
    },
    {
      "epoch": 1.0397548056894603,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 4368
    },
    {
      "epoch": 1.0399928584181397,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 4369
    },
    {
      "epoch": 1.040230911146819,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 4370
    },
    {
      "epoch": 1.0404689638754985,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 4371
    },
    {
      "epoch": 1.0407070166041779,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 4372
    },
    {
      "epoch": 1.0409450693328572,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 4373
    },
    {
      "epoch": 1.0411831220615366,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 4374
    },
    {
      "epoch": 1.041421174790216,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 4375
    },
    {
      "epoch": 1.0416592275188954,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 4376
    },
    {
      "epoch": 1.0418972802475748,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 4377
    },
    {
      "epoch": 1.0421353329762542,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 4378
    },
    {
      "epoch": 1.0423733857049335,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 4379
    },
    {
      "epoch": 1.0426114384336131,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 4380
    },
    {
      "epoch": 1.0428494911622925,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4621,
      "step": 4381
    },
    {
      "epoch": 1.043087543890972,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 4382
    },
    {
      "epoch": 1.0433255966196513,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 4383
    },
    {
      "epoch": 1.0435636493483307,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 4384
    },
    {
      "epoch": 1.04380170207701,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 4385
    },
    {
      "epoch": 1.0440397548056894,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4715,
      "step": 4386
    },
    {
      "epoch": 1.0442778075343688,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 4387
    },
    {
      "epoch": 1.0445158602630482,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 4388
    },
    {
      "epoch": 1.0447539129917276,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5277,
      "step": 4389
    },
    {
      "epoch": 1.044991965720407,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3435,
      "step": 4390
    },
    {
      "epoch": 1.0452300184490864,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 4391
    },
    {
      "epoch": 1.045468071177766,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4385,
      "step": 4392
    },
    {
      "epoch": 1.0457061239064454,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3937,
      "step": 4393
    },
    {
      "epoch": 1.0459441766351247,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 4394
    },
    {
      "epoch": 1.0461822293638041,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 4395
    },
    {
      "epoch": 1.0464202820924835,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 4396
    },
    {
      "epoch": 1.0466583348211629,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 4397
    },
    {
      "epoch": 1.0468963875498423,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 4398
    },
    {
      "epoch": 1.0471344402785216,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 4399
    },
    {
      "epoch": 1.047372493007201,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 4400
    },
    {
      "epoch": 1.047372493007201,
      "eval_runtime": 0.5215,
      "eval_samples_per_second": 1.918,
      "eval_steps_per_second": 1.918,
      "step": 4400
    },
    {
      "epoch": 1.0476105457358804,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3997,
      "step": 4401
    },
    {
      "epoch": 1.0478485984645598,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 4402
    },
    {
      "epoch": 1.0480866511932394,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 4403
    },
    {
      "epoch": 1.0483247039219188,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 4404
    },
    {
      "epoch": 1.0485627566505982,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.35,
      "step": 4405
    },
    {
      "epoch": 1.0488008093792776,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 4406
    },
    {
      "epoch": 1.049038862107957,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 4407
    },
    {
      "epoch": 1.0492769148366363,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 4408
    },
    {
      "epoch": 1.0495149675653157,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 4409
    },
    {
      "epoch": 1.049753020293995,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 4410
    },
    {
      "epoch": 1.0499910730226745,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 4411
    },
    {
      "epoch": 1.0502291257513539,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 4412
    },
    {
      "epoch": 1.0504671784800332,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 4413
    },
    {
      "epoch": 1.0507052312087128,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 4414
    },
    {
      "epoch": 1.0509432839373922,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 4415
    },
    {
      "epoch": 1.0511813366660716,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2549,
      "step": 4416
    },
    {
      "epoch": 1.051419389394751,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 4417
    },
    {
      "epoch": 1.0516574421234304,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 4418
    },
    {
      "epoch": 1.0518954948521098,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 4419
    },
    {
      "epoch": 1.0521335475807891,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.6396,
      "step": 4420
    },
    {
      "epoch": 1.0523716003094685,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5691,
      "step": 4421
    },
    {
      "epoch": 1.052609653038148,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 4422
    },
    {
      "epoch": 1.0528477057668273,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.4381,
      "step": 4423
    },
    {
      "epoch": 1.0530857584955067,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 4424
    },
    {
      "epoch": 1.053323811224186,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2511,
      "step": 4425
    },
    {
      "epoch": 1.0535618639528657,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 4426
    },
    {
      "epoch": 1.053799916681545,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 4427
    },
    {
      "epoch": 1.0540379694102244,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 4428
    },
    {
      "epoch": 1.0542760221389038,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 4429
    },
    {
      "epoch": 1.0545140748675832,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 4430
    },
    {
      "epoch": 1.0547521275962626,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 4431
    },
    {
      "epoch": 1.054990180324942,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 4432
    },
    {
      "epoch": 1.0552282330536213,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 4433
    },
    {
      "epoch": 1.0554662857823007,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 4434
    },
    {
      "epoch": 1.0557043385109801,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4271,
      "step": 4435
    },
    {
      "epoch": 1.0559423912396595,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2788,
      "step": 4436
    },
    {
      "epoch": 1.0561804439683389,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.427,
      "step": 4437
    },
    {
      "epoch": 1.0564184966970185,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 4438
    },
    {
      "epoch": 1.0566565494256979,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 4439
    },
    {
      "epoch": 1.0568946021543772,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 4440
    },
    {
      "epoch": 1.0571326548830566,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 4441
    },
    {
      "epoch": 1.057370707611736,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 4442
    },
    {
      "epoch": 1.0576087603404154,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 4443
    },
    {
      "epoch": 1.0578468130690948,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.953,
      "step": 4444
    },
    {
      "epoch": 1.0580848657977742,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4259,
      "step": 4445
    },
    {
      "epoch": 1.0583229185264535,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4026,
      "step": 4446
    },
    {
      "epoch": 1.058560971255133,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4228,
      "step": 4447
    },
    {
      "epoch": 1.0587990239838123,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4856,
      "step": 4448
    },
    {
      "epoch": 1.059037076712492,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 4449
    },
    {
      "epoch": 1.0592751294411713,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.4224,
      "step": 4450
    },
    {
      "epoch": 1.0595131821698507,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 4451
    },
    {
      "epoch": 1.05975123489853,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 4452
    },
    {
      "epoch": 1.0599892876272095,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 4453
    },
    {
      "epoch": 1.0602273403558888,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 4454
    },
    {
      "epoch": 1.0604653930845682,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 4455
    },
    {
      "epoch": 1.0607034458132476,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.3779,
      "step": 4456
    },
    {
      "epoch": 1.060941498541927,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 4457
    },
    {
      "epoch": 1.0611795512706064,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3639,
      "step": 4458
    },
    {
      "epoch": 1.0614176039992858,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 4459
    },
    {
      "epoch": 1.0616556567279654,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 4460
    },
    {
      "epoch": 1.0618937094566447,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 4461
    },
    {
      "epoch": 1.0621317621853241,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 4462
    },
    {
      "epoch": 1.0623698149140035,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 4463
    },
    {
      "epoch": 1.062607867642683,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5082,
      "step": 4464
    },
    {
      "epoch": 1.0628459203713623,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 4465
    },
    {
      "epoch": 1.0630839731000417,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5967,
      "step": 4466
    },
    {
      "epoch": 1.063322025828721,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 4467
    },
    {
      "epoch": 1.0635600785574004,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 4468
    },
    {
      "epoch": 1.0637981312860798,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 4469
    },
    {
      "epoch": 1.0640361840147592,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 4470
    },
    {
      "epoch": 1.0642742367434386,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 4471
    },
    {
      "epoch": 1.064512289472118,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 4472
    },
    {
      "epoch": 1.0647503422007976,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 4473
    },
    {
      "epoch": 1.064988394929477,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 4474
    },
    {
      "epoch": 1.0652264476581563,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 4475
    },
    {
      "epoch": 1.0654645003868357,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 4476
    },
    {
      "epoch": 1.065702553115515,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 4477
    },
    {
      "epoch": 1.0659406058441945,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 4478
    },
    {
      "epoch": 1.0661786585728739,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 4479
    },
    {
      "epoch": 1.0664167113015532,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 4480
    },
    {
      "epoch": 1.0666547640302326,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 4481
    },
    {
      "epoch": 1.066892816758912,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 4482
    },
    {
      "epoch": 1.0671308694875914,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3935,
      "step": 4483
    },
    {
      "epoch": 1.067368922216271,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 4484
    },
    {
      "epoch": 1.0676069749449504,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 4485
    },
    {
      "epoch": 1.0678450276736298,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4632,
      "step": 4486
    },
    {
      "epoch": 1.0680830804023091,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 4487
    },
    {
      "epoch": 1.0683211331309885,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 4488
    },
    {
      "epoch": 1.068559185859668,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 4489
    },
    {
      "epoch": 1.0687972385883473,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 4490
    },
    {
      "epoch": 1.0690352913170267,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 4491
    },
    {
      "epoch": 1.069273344045706,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4594,
      "step": 4492
    },
    {
      "epoch": 1.0695113967743854,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 4493
    },
    {
      "epoch": 1.0697494495030648,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 4494
    },
    {
      "epoch": 1.0699875022317444,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 4495
    },
    {
      "epoch": 1.0702255549604238,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 4496
    },
    {
      "epoch": 1.0704636076891032,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 4497
    },
    {
      "epoch": 1.0707016604177826,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 4498
    },
    {
      "epoch": 1.070939713146462,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 4499
    },
    {
      "epoch": 1.0711777658751414,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3154,
      "step": 4500
    },
    {
      "epoch": 1.0711777658751414,
      "eval_runtime": 0.5105,
      "eval_samples_per_second": 1.959,
      "eval_steps_per_second": 1.959,
      "step": 4500
    },
    {
      "epoch": 1.0714158186038207,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2922,
      "step": 4501
    },
    {
      "epoch": 1.0716538713325001,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3263,
      "step": 4502
    },
    {
      "epoch": 1.0718919240611795,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 4503
    },
    {
      "epoch": 1.0721299767898589,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 4504
    },
    {
      "epoch": 1.0723680295185383,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 4505
    },
    {
      "epoch": 1.0726060822472179,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 4506
    },
    {
      "epoch": 1.0728441349758973,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 4507
    },
    {
      "epoch": 1.0730821877045766,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 4508
    },
    {
      "epoch": 1.073320240433256,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.5805,
      "step": 4509
    },
    {
      "epoch": 1.0735582931619354,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3286,
      "step": 4510
    },
    {
      "epoch": 1.0737963458906148,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 4511
    },
    {
      "epoch": 1.0740343986192942,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 4512
    },
    {
      "epoch": 1.0742724513479736,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 4513
    },
    {
      "epoch": 1.074510504076653,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 4514
    },
    {
      "epoch": 1.0747485568053323,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 4515
    },
    {
      "epoch": 1.0749866095340117,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 4516
    },
    {
      "epoch": 1.075224662262691,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3651,
      "step": 4517
    },
    {
      "epoch": 1.0754627149913705,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 4518
    },
    {
      "epoch": 1.07570076772005,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 4519
    },
    {
      "epoch": 1.0759388204487295,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 4520
    },
    {
      "epoch": 1.0761768731774088,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 4521
    },
    {
      "epoch": 1.0764149259060882,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 4522
    },
    {
      "epoch": 1.0766529786347676,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 4523
    },
    {
      "epoch": 1.076891031363447,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 4524
    },
    {
      "epoch": 1.0771290840921264,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 4525
    },
    {
      "epoch": 1.0773671368208058,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 4526
    },
    {
      "epoch": 1.0776051895494851,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 4527
    },
    {
      "epoch": 1.0778432422781645,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 4528
    },
    {
      "epoch": 1.078081295006844,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 4529
    },
    {
      "epoch": 1.0783193477355235,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 4530
    },
    {
      "epoch": 1.078557400464203,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.4627,
      "step": 4531
    },
    {
      "epoch": 1.0787954531928823,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 4532
    },
    {
      "epoch": 1.0790335059215617,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 4533
    },
    {
      "epoch": 1.079271558650241,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 4534
    },
    {
      "epoch": 1.0795096113789204,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 4535
    },
    {
      "epoch": 1.0797476641075998,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4708,
      "step": 4536
    },
    {
      "epoch": 1.0799857168362792,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5759,
      "step": 4537
    },
    {
      "epoch": 1.0802237695649586,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 4538
    },
    {
      "epoch": 1.080461822293638,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 4539
    },
    {
      "epoch": 1.0806998750223173,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3902,
      "step": 4540
    },
    {
      "epoch": 1.080937927750997,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 4541
    },
    {
      "epoch": 1.0811759804796763,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 4542
    },
    {
      "epoch": 1.0814140332083557,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 4543
    },
    {
      "epoch": 1.081652085937035,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 4544
    },
    {
      "epoch": 1.0818901386657145,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 4545
    },
    {
      "epoch": 1.0821281913943939,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 4546
    },
    {
      "epoch": 1.0823662441230733,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 4547
    },
    {
      "epoch": 1.0826042968517526,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 4548
    },
    {
      "epoch": 1.082842349580432,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 4549
    },
    {
      "epoch": 1.0830804023091114,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 4550
    },
    {
      "epoch": 1.0833184550377908,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 4551
    },
    {
      "epoch": 1.0835565077664704,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 4552
    },
    {
      "epoch": 1.0837945604951498,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 4553
    },
    {
      "epoch": 1.0840326132238292,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3062,
      "step": 4554
    },
    {
      "epoch": 1.0842706659525085,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2526,
      "step": 4555
    },
    {
      "epoch": 1.084508718681188,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 4556
    },
    {
      "epoch": 1.0847467714098673,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 4557
    },
    {
      "epoch": 1.0849848241385467,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 4558
    },
    {
      "epoch": 1.085222876867226,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3247,
      "step": 4559
    },
    {
      "epoch": 1.0854609295959055,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4139,
      "step": 4560
    },
    {
      "epoch": 1.0856989823245848,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 4561
    },
    {
      "epoch": 1.0859370350532642,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 4562
    },
    {
      "epoch": 1.0861750877819436,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 4563
    },
    {
      "epoch": 1.086413140510623,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 4564
    },
    {
      "epoch": 1.0866511932393026,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 4565
    },
    {
      "epoch": 1.086889245967982,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 4566
    },
    {
      "epoch": 1.0871272986966614,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.5369,
      "step": 4567
    },
    {
      "epoch": 1.0873653514253407,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.331,
      "step": 4568
    },
    {
      "epoch": 1.0876034041540201,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3861,
      "step": 4569
    },
    {
      "epoch": 1.0878414568826995,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 4570
    },
    {
      "epoch": 1.088079509611379,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 4571
    },
    {
      "epoch": 1.0883175623400583,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 4572
    },
    {
      "epoch": 1.0885556150687377,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 4573
    },
    {
      "epoch": 1.088793667797417,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 4574
    },
    {
      "epoch": 1.0890317205260964,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 4575
    },
    {
      "epoch": 1.089269773254776,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 4576
    },
    {
      "epoch": 1.0895078259834554,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 4577
    },
    {
      "epoch": 1.0897458787121348,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2936,
      "step": 4578
    },
    {
      "epoch": 1.0899839314408142,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 4579
    },
    {
      "epoch": 1.0902219841694936,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 4580
    },
    {
      "epoch": 1.090460036898173,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 4581
    },
    {
      "epoch": 1.0906980896268523,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 4582
    },
    {
      "epoch": 1.0909361423555317,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 4583
    },
    {
      "epoch": 1.091174195084211,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 4584
    },
    {
      "epoch": 1.0914122478128905,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8219,
      "step": 4585
    },
    {
      "epoch": 1.0916503005415699,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.391,
      "step": 4586
    },
    {
      "epoch": 1.0918883532702495,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 4587
    },
    {
      "epoch": 1.0921264059989289,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 4588
    },
    {
      "epoch": 1.0923644587276082,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 4589
    },
    {
      "epoch": 1.0926025114562876,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 4590
    },
    {
      "epoch": 1.092840564184967,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 4591
    },
    {
      "epoch": 1.0930786169136464,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.5248,
      "step": 4592
    },
    {
      "epoch": 1.0933166696423258,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 4593
    },
    {
      "epoch": 1.0935547223710052,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 4594
    },
    {
      "epoch": 1.0937927750996845,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 4595
    },
    {
      "epoch": 1.094030827828364,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 4596
    },
    {
      "epoch": 1.0942688805570433,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3875,
      "step": 4597
    },
    {
      "epoch": 1.094506933285723,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 4598
    },
    {
      "epoch": 1.0947449860144023,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 4599
    },
    {
      "epoch": 1.0949830387430817,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 4600
    },
    {
      "epoch": 1.0949830387430817,
      "eval_runtime": 0.5157,
      "eval_samples_per_second": 1.939,
      "eval_steps_per_second": 1.939,
      "step": 4600
    },
    {
      "epoch": 1.095221091471761,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 4601
    },
    {
      "epoch": 1.0954591442004404,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 4602
    },
    {
      "epoch": 1.0956971969291198,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 4603
    },
    {
      "epoch": 1.0959352496577992,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.431,
      "step": 4604
    },
    {
      "epoch": 1.0961733023864786,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 4605
    },
    {
      "epoch": 1.096411355115158,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3398,
      "step": 4606
    },
    {
      "epoch": 1.0966494078438374,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3778,
      "step": 4607
    },
    {
      "epoch": 1.0968874605725167,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3779,
      "step": 4608
    },
    {
      "epoch": 1.0971255133011961,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 4609
    },
    {
      "epoch": 1.0973635660298755,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5429,
      "step": 4610
    },
    {
      "epoch": 1.097601618758555,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4425,
      "step": 4611
    },
    {
      "epoch": 1.0978396714872345,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 4612
    },
    {
      "epoch": 1.0980777242159139,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 4613
    },
    {
      "epoch": 1.0983157769445933,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 4614
    },
    {
      "epoch": 1.0985538296732726,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2439,
      "step": 4615
    },
    {
      "epoch": 1.098791882401952,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 4616
    },
    {
      "epoch": 1.0990299351306314,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 4617
    },
    {
      "epoch": 1.0992679878593108,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 4618
    },
    {
      "epoch": 1.0995060405879902,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 4619
    },
    {
      "epoch": 1.0997440933166696,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.499,
      "step": 4620
    },
    {
      "epoch": 1.099982146045349,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 4621
    },
    {
      "epoch": 1.1002201987740285,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 4622
    },
    {
      "epoch": 1.100458251502708,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 4623
    },
    {
      "epoch": 1.1006963042313873,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 4624
    },
    {
      "epoch": 1.1009343569600667,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 4625
    },
    {
      "epoch": 1.101172409688746,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 4626
    },
    {
      "epoch": 1.1014104624174255,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.5009,
      "step": 4627
    },
    {
      "epoch": 1.1016485151461048,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 4628
    },
    {
      "epoch": 1.1018865678747842,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 4629
    },
    {
      "epoch": 1.1021246206034636,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3708,
      "step": 4630
    },
    {
      "epoch": 1.102362673332143,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.4352,
      "step": 4631
    },
    {
      "epoch": 1.1026007260608224,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3031,
      "step": 4632
    },
    {
      "epoch": 1.102838778789502,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 4633
    },
    {
      "epoch": 1.1030768315181814,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 4634
    },
    {
      "epoch": 1.1033148842468607,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 4635
    },
    {
      "epoch": 1.1035529369755401,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 4636
    },
    {
      "epoch": 1.1037909897042195,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 4637
    },
    {
      "epoch": 1.104029042432899,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 4638
    },
    {
      "epoch": 1.1042670951615783,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 4639
    },
    {
      "epoch": 1.1045051478902577,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 4640
    },
    {
      "epoch": 1.104743200618937,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 4641
    },
    {
      "epoch": 1.1049812533476164,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 4642
    },
    {
      "epoch": 1.1052193060762958,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 4643
    },
    {
      "epoch": 1.1054573588049752,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 4644
    },
    {
      "epoch": 1.1056954115336548,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 4645
    },
    {
      "epoch": 1.1059334642623342,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 4646
    },
    {
      "epoch": 1.1061715169910136,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 4647
    },
    {
      "epoch": 1.106409569719693,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3918,
      "step": 4648
    },
    {
      "epoch": 1.1066476224483723,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3265,
      "step": 4649
    },
    {
      "epoch": 1.1068856751770517,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 4650
    },
    {
      "epoch": 1.107123727905731,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4472,
      "step": 4651
    },
    {
      "epoch": 1.1073617806344105,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 4652
    },
    {
      "epoch": 1.1075998333630899,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 4653
    },
    {
      "epoch": 1.1078378860917693,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 4654
    },
    {
      "epoch": 1.1080759388204486,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 4655
    },
    {
      "epoch": 1.108313991549128,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 4656
    },
    {
      "epoch": 1.1085520442778076,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 4657
    },
    {
      "epoch": 1.108790097006487,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 4658
    },
    {
      "epoch": 1.1090281497351664,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 4659
    },
    {
      "epoch": 1.1092662024638458,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 4660
    },
    {
      "epoch": 1.1095042551925252,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 4661
    },
    {
      "epoch": 1.1097423079212045,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 4662
    },
    {
      "epoch": 1.109980360649884,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 4663
    },
    {
      "epoch": 1.1102184133785633,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 4664
    },
    {
      "epoch": 1.1104564661072427,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 4665
    },
    {
      "epoch": 1.110694518835922,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 4666
    },
    {
      "epoch": 1.1109325715646015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 4667
    },
    {
      "epoch": 1.111170624293281,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 4668
    },
    {
      "epoch": 1.1114086770219604,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 4669
    },
    {
      "epoch": 1.1116467297506398,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 4670
    },
    {
      "epoch": 1.1118847824793192,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 4671
    },
    {
      "epoch": 1.1121228352079986,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 4672
    },
    {
      "epoch": 1.112360887936678,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.5161,
      "step": 4673
    },
    {
      "epoch": 1.1125989406653574,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3375,
      "step": 4674
    },
    {
      "epoch": 1.1128369933940367,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 4675
    },
    {
      "epoch": 1.1130750461227161,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4935,
      "step": 4676
    },
    {
      "epoch": 1.1133130988513955,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 4677
    },
    {
      "epoch": 1.113551151580075,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 4678
    },
    {
      "epoch": 1.1137892043087545,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4033,
      "step": 4679
    },
    {
      "epoch": 1.1140272570374339,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 4680
    },
    {
      "epoch": 1.1142653097661133,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 4681
    },
    {
      "epoch": 1.1145033624947926,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 4682
    },
    {
      "epoch": 1.114741415223472,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 4683
    },
    {
      "epoch": 1.1149794679521514,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 4684
    },
    {
      "epoch": 1.1152175206808308,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 4685
    },
    {
      "epoch": 1.1154555734095102,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 4686
    },
    {
      "epoch": 1.1156936261381896,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 4687
    },
    {
      "epoch": 1.115931678866869,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.6009,
      "step": 4688
    },
    {
      "epoch": 1.1161697315955483,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.346,
      "step": 4689
    },
    {
      "epoch": 1.1164077843242277,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 4690
    },
    {
      "epoch": 1.1166458370529073,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3939,
      "step": 4691
    },
    {
      "epoch": 1.1168838897815867,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 4692
    },
    {
      "epoch": 1.117121942510266,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 4693
    },
    {
      "epoch": 1.1173599952389455,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 4694
    },
    {
      "epoch": 1.1175980479676249,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.6001,
      "step": 4695
    },
    {
      "epoch": 1.1178361006963042,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 4696
    },
    {
      "epoch": 1.1180741534249836,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 4697
    },
    {
      "epoch": 1.118312206153663,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 4698
    },
    {
      "epoch": 1.1185502588823424,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 4699
    },
    {
      "epoch": 1.1187883116110218,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 4700
    },
    {
      "epoch": 1.1187883116110218,
      "eval_runtime": 0.5181,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 4700
    },
    {
      "epoch": 1.1190263643397012,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 4701
    },
    {
      "epoch": 1.1192644170683805,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4367,
      "step": 4702
    },
    {
      "epoch": 1.1195024697970601,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3986,
      "step": 4703
    },
    {
      "epoch": 1.1197405225257395,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 4704
    },
    {
      "epoch": 1.119978575254419,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 4705
    },
    {
      "epoch": 1.1202166279830983,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2674,
      "step": 4706
    },
    {
      "epoch": 1.1204546807117777,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 4707
    },
    {
      "epoch": 1.120692733440457,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3566,
      "step": 4708
    },
    {
      "epoch": 1.1209307861691364,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 4709
    },
    {
      "epoch": 1.1211688388978158,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 4710
    },
    {
      "epoch": 1.1214068916264952,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 4711
    },
    {
      "epoch": 1.1216449443551746,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 4712
    },
    {
      "epoch": 1.121882997083854,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.4625,
      "step": 4713
    },
    {
      "epoch": 1.1221210498125336,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 4714
    },
    {
      "epoch": 1.122359102541213,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 4715
    },
    {
      "epoch": 1.1225971552698923,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5833,
      "step": 4716
    },
    {
      "epoch": 1.1228352079985717,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 4717
    },
    {
      "epoch": 1.123073260727251,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 4718
    },
    {
      "epoch": 1.1233113134559305,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5212,
      "step": 4719
    },
    {
      "epoch": 1.1235493661846099,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 4720
    },
    {
      "epoch": 1.1237874189132893,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 4721
    },
    {
      "epoch": 1.1240254716419686,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3297,
      "step": 4722
    },
    {
      "epoch": 1.124263524370648,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 4723
    },
    {
      "epoch": 1.1245015770993274,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 4724
    },
    {
      "epoch": 1.124739629828007,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 4725
    },
    {
      "epoch": 1.1249776825566864,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 4726
    },
    {
      "epoch": 1.1252157352853658,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 4727
    },
    {
      "epoch": 1.1254537880140452,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2621,
      "step": 4728
    },
    {
      "epoch": 1.1256918407427245,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 4729
    },
    {
      "epoch": 1.125929893471404,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 4730
    },
    {
      "epoch": 1.1261679462000833,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 4731
    },
    {
      "epoch": 1.1264059989287627,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 4732
    },
    {
      "epoch": 1.126644051657442,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2997,
      "step": 4733
    },
    {
      "epoch": 1.1268821043861215,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 4734
    },
    {
      "epoch": 1.1271201571148008,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2428,
      "step": 4735
    },
    {
      "epoch": 1.1273582098434805,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4379,
      "step": 4736
    },
    {
      "epoch": 1.1275962625721596,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 4737
    },
    {
      "epoch": 1.1278343153008392,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 4738
    },
    {
      "epoch": 1.1280723680295186,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 4739
    },
    {
      "epoch": 1.128310420758198,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4808,
      "step": 4740
    },
    {
      "epoch": 1.1285484734868774,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4063,
      "step": 4741
    },
    {
      "epoch": 1.1287865262155568,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.238,
      "step": 4742
    },
    {
      "epoch": 1.1290245789442361,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 4743
    },
    {
      "epoch": 1.1292626316729155,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 4744
    },
    {
      "epoch": 1.129500684401595,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 4745
    },
    {
      "epoch": 1.1297387371302743,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 4746
    },
    {
      "epoch": 1.1299767898589537,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 4747
    },
    {
      "epoch": 1.130214842587633,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 4748
    },
    {
      "epoch": 1.1304528953163127,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 4749
    },
    {
      "epoch": 1.130690948044992,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 4750
    },
    {
      "epoch": 1.1309290007736714,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 4751
    },
    {
      "epoch": 1.1311670535023508,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 4752
    },
    {
      "epoch": 1.1314051062310302,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1747,
      "step": 4753
    },
    {
      "epoch": 1.1316431589597096,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 4754
    },
    {
      "epoch": 1.131881211688389,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 4755
    },
    {
      "epoch": 1.1321192644170683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 4756
    },
    {
      "epoch": 1.1323573171457477,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 4757
    },
    {
      "epoch": 1.132595369874427,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 4758
    },
    {
      "epoch": 1.1328334226031065,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 4759
    },
    {
      "epoch": 1.133071475331786,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 4760
    },
    {
      "epoch": 1.1333095280604655,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 4761
    },
    {
      "epoch": 1.1335475807891449,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 4762
    },
    {
      "epoch": 1.1337856335178242,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.6221,
      "step": 4763
    },
    {
      "epoch": 1.1340236862465036,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3553,
      "step": 4764
    },
    {
      "epoch": 1.134261738975183,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4048,
      "step": 4765
    },
    {
      "epoch": 1.1344997917038624,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 4766
    },
    {
      "epoch": 1.1347378444325418,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3661,
      "step": 4767
    },
    {
      "epoch": 1.1349758971612212,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 4768
    },
    {
      "epoch": 1.1352139498899005,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 4769
    },
    {
      "epoch": 1.13545200261858,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3859,
      "step": 4770
    },
    {
      "epoch": 1.1356900553472595,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 4771
    },
    {
      "epoch": 1.135928108075939,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.6009,
      "step": 4772
    },
    {
      "epoch": 1.1361661608046183,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 4773
    },
    {
      "epoch": 1.1364042135332977,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 4774
    },
    {
      "epoch": 1.136642266261977,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 4775
    },
    {
      "epoch": 1.1368803189906564,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8579,
      "step": 4776
    },
    {
      "epoch": 1.1371183717193358,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 4777
    },
    {
      "epoch": 1.1373564244480152,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2762,
      "step": 4778
    },
    {
      "epoch": 1.1375944771766946,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 4779
    },
    {
      "epoch": 1.137832529905374,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 4780
    },
    {
      "epoch": 1.1380705826340534,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 4781
    },
    {
      "epoch": 1.138308635362733,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 4782
    },
    {
      "epoch": 1.1385466880914121,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 4783
    },
    {
      "epoch": 1.1387847408200917,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 4784
    },
    {
      "epoch": 1.1390227935487711,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 4785
    },
    {
      "epoch": 1.1392608462774505,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1506,
      "step": 4786
    },
    {
      "epoch": 1.1394988990061299,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 4787
    },
    {
      "epoch": 1.1397369517348093,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 4788
    },
    {
      "epoch": 1.1399750044634887,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 4789
    },
    {
      "epoch": 1.140213057192168,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 4790
    },
    {
      "epoch": 1.1404511099208474,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 4791
    },
    {
      "epoch": 1.1406891626495268,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4313,
      "step": 4792
    },
    {
      "epoch": 1.1409272153782062,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 4793
    },
    {
      "epoch": 1.1411652681068856,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 4794
    },
    {
      "epoch": 1.1414033208355652,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 4795
    },
    {
      "epoch": 1.1416413735642446,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 4796
    },
    {
      "epoch": 1.141879426292924,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 4797
    },
    {
      "epoch": 1.1421174790216033,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 4798
    },
    {
      "epoch": 1.1423555317502827,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 4799
    },
    {
      "epoch": 1.142593584478962,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 4800
    },
    {
      "epoch": 1.142593584478962,
      "eval_runtime": 0.5294,
      "eval_samples_per_second": 1.889,
      "eval_steps_per_second": 1.889,
      "step": 4800
    },
    {
      "epoch": 1.1428316372076415,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 4801
    },
    {
      "epoch": 1.1430696899363209,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 4802
    },
    {
      "epoch": 1.1433077426650002,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.7029,
      "step": 4803
    },
    {
      "epoch": 1.1435457953936796,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2643,
      "step": 4804
    },
    {
      "epoch": 1.143783848122359,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 4805
    },
    {
      "epoch": 1.1440219008510386,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 4806
    },
    {
      "epoch": 1.144259953579718,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 4807
    },
    {
      "epoch": 1.1444980063083974,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 4808
    },
    {
      "epoch": 1.1447360590370768,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 4809
    },
    {
      "epoch": 1.1449741117657561,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 4810
    },
    {
      "epoch": 1.1452121644944355,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 4811
    },
    {
      "epoch": 1.145450217223115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1899,
      "step": 4812
    },
    {
      "epoch": 1.1456882699517943,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 4813
    },
    {
      "epoch": 1.1459263226804737,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 4814
    },
    {
      "epoch": 1.146164375409153,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 4815
    },
    {
      "epoch": 1.1464024281378324,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 4816
    },
    {
      "epoch": 1.146640480866512,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2438,
      "step": 4817
    },
    {
      "epoch": 1.1468785335951914,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 4818
    },
    {
      "epoch": 1.1471165863238708,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 4819
    },
    {
      "epoch": 1.1473546390525502,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 4820
    },
    {
      "epoch": 1.1475926917812296,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 4821
    },
    {
      "epoch": 1.147830744509909,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 4822
    },
    {
      "epoch": 1.1480687972385883,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 4823
    },
    {
      "epoch": 1.1483068499672677,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 4824
    },
    {
      "epoch": 1.1485449026959471,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 4825
    },
    {
      "epoch": 1.1487829554246265,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 4826
    },
    {
      "epoch": 1.1490210081533059,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 4827
    },
    {
      "epoch": 1.1492590608819855,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 4828
    },
    {
      "epoch": 1.1494971136106646,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 4829
    },
    {
      "epoch": 1.1497351663393443,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 4830
    },
    {
      "epoch": 1.1499732190680236,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 4831
    },
    {
      "epoch": 1.150211271796703,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 4832
    },
    {
      "epoch": 1.1504493245253824,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 4833
    },
    {
      "epoch": 1.1506873772540618,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 4834
    },
    {
      "epoch": 1.1509254299827412,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 4835
    },
    {
      "epoch": 1.1511634827114205,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 4836
    },
    {
      "epoch": 1.1514015354401,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 4837
    },
    {
      "epoch": 1.1516395881687793,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 4838
    },
    {
      "epoch": 1.1518776408974587,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 4839
    },
    {
      "epoch": 1.152115693626138,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 4840
    },
    {
      "epoch": 1.1523537463548177,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 4841
    },
    {
      "epoch": 1.152591799083497,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 4842
    },
    {
      "epoch": 1.1528298518121765,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4652,
      "step": 4843
    },
    {
      "epoch": 1.1530679045408558,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 4844
    },
    {
      "epoch": 1.1533059572695352,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 4845
    },
    {
      "epoch": 1.1535440099982146,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 4846
    },
    {
      "epoch": 1.153782062726894,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 4847
    },
    {
      "epoch": 1.1540201154555734,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 4848
    },
    {
      "epoch": 1.1542581681842528,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.6634,
      "step": 4849
    },
    {
      "epoch": 1.1544962209129321,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 4850
    },
    {
      "epoch": 1.1547342736416115,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 4851
    },
    {
      "epoch": 1.1549723263702911,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.5255,
      "step": 4852
    },
    {
      "epoch": 1.1552103790989705,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 4853
    },
    {
      "epoch": 1.15544843182765,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 4854
    },
    {
      "epoch": 1.1556864845563293,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 4855
    },
    {
      "epoch": 1.1559245372850087,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 4856
    },
    {
      "epoch": 1.156162590013688,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 4857
    },
    {
      "epoch": 1.1564006427423674,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 4858
    },
    {
      "epoch": 1.1566386954710468,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 4859
    },
    {
      "epoch": 1.1568767481997262,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 4860
    },
    {
      "epoch": 1.1571148009284056,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 4861
    },
    {
      "epoch": 1.157352853657085,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 4862
    },
    {
      "epoch": 1.1575909063857646,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 4863
    },
    {
      "epoch": 1.157828959114444,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4497,
      "step": 4864
    },
    {
      "epoch": 1.1580670118431233,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 4865
    },
    {
      "epoch": 1.1583050645718027,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 4866
    },
    {
      "epoch": 1.158543117300482,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 4867
    },
    {
      "epoch": 1.1587811700291615,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 4868
    },
    {
      "epoch": 1.1590192227578409,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 4869
    },
    {
      "epoch": 1.1592572754865202,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 4870
    },
    {
      "epoch": 1.1594953282151996,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 4871
    },
    {
      "epoch": 1.159733380943879,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 4872
    },
    {
      "epoch": 1.1599714336725584,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 4873
    },
    {
      "epoch": 1.1602094864012378,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 4874
    },
    {
      "epoch": 1.1604475391299172,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3299,
      "step": 4875
    },
    {
      "epoch": 1.1606855918585968,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 4876
    },
    {
      "epoch": 1.1609236445872761,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 4877
    },
    {
      "epoch": 1.1611616973159555,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 4878
    },
    {
      "epoch": 1.161399750044635,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 4879
    },
    {
      "epoch": 1.1616378027733143,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 4880
    },
    {
      "epoch": 1.1618758555019937,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 4881
    },
    {
      "epoch": 1.162113908230673,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 4882
    },
    {
      "epoch": 1.1623519609593524,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 4883
    },
    {
      "epoch": 1.1625900136880318,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 4884
    },
    {
      "epoch": 1.1628280664167112,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 4885
    },
    {
      "epoch": 1.1630661191453906,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 4886
    },
    {
      "epoch": 1.1633041718740702,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 4887
    },
    {
      "epoch": 1.1635422246027496,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 4888
    },
    {
      "epoch": 1.163780277331429,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 4889
    },
    {
      "epoch": 1.1640183300601084,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 4890
    },
    {
      "epoch": 1.1642563827887877,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2695,
      "step": 4891
    },
    {
      "epoch": 1.1644944355174671,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 4892
    },
    {
      "epoch": 1.1647324882461465,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 4893
    },
    {
      "epoch": 1.1649705409748259,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4761,
      "step": 4894
    },
    {
      "epoch": 1.1652085937035053,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 4895
    },
    {
      "epoch": 1.1654466464321847,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2663,
      "step": 4896
    },
    {
      "epoch": 1.165684699160864,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 4897
    },
    {
      "epoch": 1.1659227518895436,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 4898
    },
    {
      "epoch": 1.166160804618223,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 4899
    },
    {
      "epoch": 1.1663988573469024,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 4900
    },
    {
      "epoch": 1.1663988573469024,
      "eval_runtime": 0.5143,
      "eval_samples_per_second": 1.944,
      "eval_steps_per_second": 1.944,
      "step": 4900
    },
    {
      "epoch": 1.1666369100755818,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 4901
    },
    {
      "epoch": 1.1668749628042612,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 4902
    },
    {
      "epoch": 1.1671130155329406,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2038,
      "step": 4903
    },
    {
      "epoch": 1.16735106826162,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 4904
    },
    {
      "epoch": 1.1675891209902993,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 4905
    },
    {
      "epoch": 1.1678271737189787,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 4906
    },
    {
      "epoch": 1.168065226447658,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 4907
    },
    {
      "epoch": 1.1683032791763375,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 4908
    },
    {
      "epoch": 1.168541331905017,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 4909
    },
    {
      "epoch": 1.1687793846336962,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 4910
    },
    {
      "epoch": 1.1690174373623758,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 4911
    },
    {
      "epoch": 1.1692554900910552,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 4912
    },
    {
      "epoch": 1.1694935428197346,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 4913
    },
    {
      "epoch": 1.169731595548414,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 4914
    },
    {
      "epoch": 1.1699696482770934,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 4915
    },
    {
      "epoch": 1.1702077010057728,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 4916
    },
    {
      "epoch": 1.1704457537344521,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 4917
    },
    {
      "epoch": 1.1706838064631315,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 4918
    },
    {
      "epoch": 1.170921859191811,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 4919
    },
    {
      "epoch": 1.1711599119204903,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.4934,
      "step": 4920
    },
    {
      "epoch": 1.1713979646491697,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 4921
    },
    {
      "epoch": 1.1716360173778493,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 4922
    },
    {
      "epoch": 1.1718740701065287,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3778,
      "step": 4923
    },
    {
      "epoch": 1.172112122835208,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 4924
    },
    {
      "epoch": 1.1723501755638874,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 4925
    },
    {
      "epoch": 1.1725882282925668,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 4926
    },
    {
      "epoch": 1.1728262810212462,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 4927
    },
    {
      "epoch": 1.1730643337499256,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4921,
      "step": 4928
    },
    {
      "epoch": 1.173302386478605,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3563,
      "step": 4929
    },
    {
      "epoch": 1.1735404392072843,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 4930
    },
    {
      "epoch": 1.1737784919359637,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 4931
    },
    {
      "epoch": 1.1740165446646431,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 4932
    },
    {
      "epoch": 1.1742545973933227,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 4933
    },
    {
      "epoch": 1.174492650122002,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 4934
    },
    {
      "epoch": 1.1747307028506815,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 4935
    },
    {
      "epoch": 1.1749687555793609,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 4936
    },
    {
      "epoch": 1.1752068083080403,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 4937
    },
    {
      "epoch": 1.1754448610367196,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 4938
    },
    {
      "epoch": 1.175682913765399,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 4939
    },
    {
      "epoch": 1.1759209664940784,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 4940
    },
    {
      "epoch": 1.1761590192227578,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4598,
      "step": 4941
    },
    {
      "epoch": 1.1763970719514372,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 4942
    },
    {
      "epoch": 1.1766351246801166,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 4943
    },
    {
      "epoch": 1.1768731774087962,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 4944
    },
    {
      "epoch": 1.1771112301374755,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 4945
    },
    {
      "epoch": 1.177349282866155,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 4946
    },
    {
      "epoch": 1.1775873355948343,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 4947
    },
    {
      "epoch": 1.1778253883235137,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 4948
    },
    {
      "epoch": 1.178063441052193,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 4949
    },
    {
      "epoch": 1.1783014937808725,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 4950
    },
    {
      "epoch": 1.1785395465095518,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8173,
      "step": 4951
    },
    {
      "epoch": 1.1787775992382312,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 4952
    },
    {
      "epoch": 1.1790156519669106,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.6456,
      "step": 4953
    },
    {
      "epoch": 1.17925370469559,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 4954
    },
    {
      "epoch": 1.1794917574242696,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 4955
    },
    {
      "epoch": 1.1797298101529488,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 4956
    },
    {
      "epoch": 1.1799678628816284,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 4957
    },
    {
      "epoch": 1.1802059156103077,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 4958
    },
    {
      "epoch": 1.1804439683389871,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 4959
    },
    {
      "epoch": 1.1806820210676665,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 4960
    },
    {
      "epoch": 1.180920073796346,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 4961
    },
    {
      "epoch": 1.1811581265250253,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 4962
    },
    {
      "epoch": 1.1813961792537047,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 4963
    },
    {
      "epoch": 1.181634231982384,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 4964
    },
    {
      "epoch": 1.1818722847110634,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 4965
    },
    {
      "epoch": 1.1821103374397428,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 4966
    },
    {
      "epoch": 1.1823483901684222,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 4967
    },
    {
      "epoch": 1.1825864428971018,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 4968
    },
    {
      "epoch": 1.1828244956257812,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3705,
      "step": 4969
    },
    {
      "epoch": 1.1830625483544606,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 4970
    },
    {
      "epoch": 1.18330060108314,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 4971
    },
    {
      "epoch": 1.1835386538118193,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 4972
    },
    {
      "epoch": 1.1837767065404987,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 4973
    },
    {
      "epoch": 1.184014759269178,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 4974
    },
    {
      "epoch": 1.1842528119978575,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3787,
      "step": 4975
    },
    {
      "epoch": 1.1844908647265369,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 4976
    },
    {
      "epoch": 1.1847289174552162,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 4977
    },
    {
      "epoch": 1.1849669701838956,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3837,
      "step": 4978
    },
    {
      "epoch": 1.1852050229125752,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5959,
      "step": 4979
    },
    {
      "epoch": 1.1854430756412546,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 4980
    },
    {
      "epoch": 1.185681128369934,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 4981
    },
    {
      "epoch": 1.1859191810986134,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 4982
    },
    {
      "epoch": 1.1861572338272928,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2037,
      "step": 4983
    },
    {
      "epoch": 1.1863952865559722,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 4984
    },
    {
      "epoch": 1.1866333392846515,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4428,
      "step": 4985
    },
    {
      "epoch": 1.186871392013331,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 4986
    },
    {
      "epoch": 1.1871094447420103,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2744,
      "step": 4987
    },
    {
      "epoch": 1.1873474974706897,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 4988
    },
    {
      "epoch": 1.187585550199369,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 4989
    },
    {
      "epoch": 1.1878236029280487,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 4990
    },
    {
      "epoch": 1.188061655656728,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.5055,
      "step": 4991
    },
    {
      "epoch": 1.1882997083854074,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.4206,
      "step": 4992
    },
    {
      "epoch": 1.1885377611140868,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 4993
    },
    {
      "epoch": 1.1887758138427662,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 4994
    },
    {
      "epoch": 1.1890138665714456,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 4995
    },
    {
      "epoch": 1.189251919300125,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 4996
    },
    {
      "epoch": 1.1894899720288044,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4005,
      "step": 4997
    },
    {
      "epoch": 1.1897280247574837,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3943,
      "step": 4998
    },
    {
      "epoch": 1.1899660774861631,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4044,
      "step": 4999
    },
    {
      "epoch": 1.1902041302148425,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 5000
    },
    {
      "epoch": 1.1902041302148425,
      "eval_runtime": 0.5084,
      "eval_samples_per_second": 1.967,
      "eval_steps_per_second": 1.967,
      "step": 5000
    },
    {
      "epoch": 1.190442182943522,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 5001
    },
    {
      "epoch": 1.1906802356722013,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 5002
    },
    {
      "epoch": 1.1909182884008809,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 5003
    },
    {
      "epoch": 1.1911563411295603,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4683,
      "step": 5004
    },
    {
      "epoch": 1.1913943938582396,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 5005
    },
    {
      "epoch": 1.191632446586919,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 5006
    },
    {
      "epoch": 1.1918704993155984,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 5007
    },
    {
      "epoch": 1.1921085520442778,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 5008
    },
    {
      "epoch": 1.1923466047729572,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 5009
    },
    {
      "epoch": 1.1925846575016366,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 5010
    },
    {
      "epoch": 1.192822710230316,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 5011
    },
    {
      "epoch": 1.1930607629589953,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.5634,
      "step": 5012
    },
    {
      "epoch": 1.1932988156876747,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 5013
    },
    {
      "epoch": 1.1935368684163543,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 5014
    },
    {
      "epoch": 1.1937749211450337,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 5015
    },
    {
      "epoch": 1.194012973873713,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 5016
    },
    {
      "epoch": 1.1942510266023925,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 5017
    },
    {
      "epoch": 1.1944890793310718,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 5018
    },
    {
      "epoch": 1.1947271320597512,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 5019
    },
    {
      "epoch": 1.1949651847884306,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 5020
    },
    {
      "epoch": 1.19520323751711,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 5021
    },
    {
      "epoch": 1.1954412902457894,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 5022
    },
    {
      "epoch": 1.1956793429744688,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 5023
    },
    {
      "epoch": 1.1959173957031481,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 5024
    },
    {
      "epoch": 1.1961554484318278,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 5025
    },
    {
      "epoch": 1.1963935011605071,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3999,
      "step": 5026
    },
    {
      "epoch": 1.1966315538891865,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 5027
    },
    {
      "epoch": 1.196869606617866,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 5028
    },
    {
      "epoch": 1.1971076593465453,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 5029
    },
    {
      "epoch": 1.1973457120752247,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 5030
    },
    {
      "epoch": 1.197583764803904,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 5031
    },
    {
      "epoch": 1.1978218175325834,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 5032
    },
    {
      "epoch": 1.1980598702612628,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 5033
    },
    {
      "epoch": 1.1982979229899422,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 5034
    },
    {
      "epoch": 1.1985359757186216,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 5035
    },
    {
      "epoch": 1.1987740284473012,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 5036
    },
    {
      "epoch": 1.1990120811759806,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 5037
    },
    {
      "epoch": 1.19925013390466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 5038
    },
    {
      "epoch": 1.1994881866333393,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 5039
    },
    {
      "epoch": 1.1997262393620187,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 5040
    },
    {
      "epoch": 1.199964292090698,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 5041
    },
    {
      "epoch": 1.2002023448193775,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 5042
    },
    {
      "epoch": 1.2004403975480569,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 5043
    },
    {
      "epoch": 1.2006784502767363,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.532,
      "step": 5044
    },
    {
      "epoch": 1.2009165030054156,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 5045
    },
    {
      "epoch": 1.201154555734095,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 5046
    },
    {
      "epoch": 1.2013926084627746,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 5047
    },
    {
      "epoch": 1.2016306611914538,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 5048
    },
    {
      "epoch": 1.2018687139201334,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 5049
    },
    {
      "epoch": 1.2021067666488128,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 5050
    },
    {
      "epoch": 1.2023448193774922,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 5051
    },
    {
      "epoch": 1.2025828721061715,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 5052
    },
    {
      "epoch": 1.202820924834851,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.29,
      "step": 5053
    },
    {
      "epoch": 1.2030589775635303,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 5054
    },
    {
      "epoch": 1.2032970302922097,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3232,
      "step": 5055
    },
    {
      "epoch": 1.203535083020889,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 5056
    },
    {
      "epoch": 1.2037731357495685,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 5057
    },
    {
      "epoch": 1.2040111884782478,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 5058
    },
    {
      "epoch": 1.2042492412069272,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 5059
    },
    {
      "epoch": 1.2044872939356068,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3587,
      "step": 5060
    },
    {
      "epoch": 1.2047253466642862,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 5061
    },
    {
      "epoch": 1.2049633993929656,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 5062
    },
    {
      "epoch": 1.205201452121645,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 5063
    },
    {
      "epoch": 1.2054395048503244,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 5064
    },
    {
      "epoch": 1.2056775575790037,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 5065
    },
    {
      "epoch": 1.2059156103076831,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 5066
    },
    {
      "epoch": 1.2061536630363625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 5067
    },
    {
      "epoch": 1.206391715765042,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 5068
    },
    {
      "epoch": 1.2066297684937213,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 5069
    },
    {
      "epoch": 1.2068678212224007,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 5070
    },
    {
      "epoch": 1.2071058739510803,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2828,
      "step": 5071
    },
    {
      "epoch": 1.2073439266797596,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3821,
      "step": 5072
    },
    {
      "epoch": 1.207581979408439,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 5073
    },
    {
      "epoch": 1.2078200321371184,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 5074
    },
    {
      "epoch": 1.2080580848657978,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 5075
    },
    {
      "epoch": 1.2082961375944772,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 5076
    },
    {
      "epoch": 1.2085341903231566,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 5077
    },
    {
      "epoch": 1.208772243051836,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.4042,
      "step": 5078
    },
    {
      "epoch": 1.2090102957805153,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 5079
    },
    {
      "epoch": 1.2092483485091947,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 5080
    },
    {
      "epoch": 1.209486401237874,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 5081
    },
    {
      "epoch": 1.2097244539665537,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 5082
    },
    {
      "epoch": 1.209962506695233,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 5083
    },
    {
      "epoch": 1.2102005594239125,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 5084
    },
    {
      "epoch": 1.2104386121525919,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 5085
    },
    {
      "epoch": 1.2106766648812712,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 5086
    },
    {
      "epoch": 1.2109147176099506,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 5087
    },
    {
      "epoch": 1.21115277033863,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4662,
      "step": 5088
    },
    {
      "epoch": 1.2113908230673094,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 5089
    },
    {
      "epoch": 1.2116288757959888,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 5090
    },
    {
      "epoch": 1.2118669285246682,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 5091
    },
    {
      "epoch": 1.2121049812533475,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 5092
    },
    {
      "epoch": 1.2123430339820271,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3829,
      "step": 5093
    },
    {
      "epoch": 1.2125810867107063,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 5094
    },
    {
      "epoch": 1.212819139439386,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 5095
    },
    {
      "epoch": 1.2130571921680653,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 5096
    },
    {
      "epoch": 1.2132952448967447,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 5097
    },
    {
      "epoch": 1.213533297625424,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 5098
    },
    {
      "epoch": 1.2137713503541034,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 5099
    },
    {
      "epoch": 1.2140094030827828,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 5100
    },
    {
      "epoch": 1.2140094030827828,
      "eval_runtime": 0.514,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 5100
    },
    {
      "epoch": 1.2142474558114622,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 5101
    },
    {
      "epoch": 1.2144855085401416,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 5102
    },
    {
      "epoch": 1.214723561268821,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 5103
    },
    {
      "epoch": 1.2149616139975004,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 5104
    },
    {
      "epoch": 1.2151996667261797,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 5105
    },
    {
      "epoch": 1.2154377194548593,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 5106
    },
    {
      "epoch": 1.2156757721835387,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 5107
    },
    {
      "epoch": 1.215913824912218,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 5108
    },
    {
      "epoch": 1.2161518776408975,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 5109
    },
    {
      "epoch": 1.2163899303695769,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5202,
      "step": 5110
    },
    {
      "epoch": 1.2166279830982563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 5111
    },
    {
      "epoch": 1.2168660358269356,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.4606,
      "step": 5112
    },
    {
      "epoch": 1.217104088555615,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 5113
    },
    {
      "epoch": 1.2173421412842944,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 5114
    },
    {
      "epoch": 1.2175801940129738,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 5115
    },
    {
      "epoch": 1.2178182467416532,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 5116
    },
    {
      "epoch": 1.2180562994703328,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 5117
    },
    {
      "epoch": 1.2182943521990122,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 5118
    },
    {
      "epoch": 1.2185324049276915,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 5119
    },
    {
      "epoch": 1.218770457656371,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4899,
      "step": 5120
    },
    {
      "epoch": 1.2190085103850503,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 5121
    },
    {
      "epoch": 1.2192465631137297,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 5122
    },
    {
      "epoch": 1.219484615842409,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.612,
      "step": 5123
    },
    {
      "epoch": 1.2197226685710885,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 5124
    },
    {
      "epoch": 1.2199607212997678,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 5125
    },
    {
      "epoch": 1.2201987740284472,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 5126
    },
    {
      "epoch": 1.2204368267571266,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4923,
      "step": 5127
    },
    {
      "epoch": 1.2206748794858062,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 5128
    },
    {
      "epoch": 1.2209129322144856,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 5129
    },
    {
      "epoch": 1.221150984943165,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4677,
      "step": 5130
    },
    {
      "epoch": 1.2213890376718444,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 5131
    },
    {
      "epoch": 1.2216270904005238,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 5132
    },
    {
      "epoch": 1.2218651431292031,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 5133
    },
    {
      "epoch": 1.2221031958578825,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 5134
    },
    {
      "epoch": 1.222341248586562,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.553,
      "step": 5135
    },
    {
      "epoch": 1.2225793013152413,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 5136
    },
    {
      "epoch": 1.2228173540439207,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 5137
    },
    {
      "epoch": 1.2230554067726,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 5138
    },
    {
      "epoch": 1.2232934595012797,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 5139
    },
    {
      "epoch": 1.2235315122299588,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3814,
      "step": 5140
    },
    {
      "epoch": 1.2237695649586384,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 5141
    },
    {
      "epoch": 1.2240076176873178,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2835,
      "step": 5142
    },
    {
      "epoch": 1.2242456704159972,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 5143
    },
    {
      "epoch": 1.2244837231446766,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 5144
    },
    {
      "epoch": 1.224721775873356,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 5145
    },
    {
      "epoch": 1.2249598286020353,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 5146
    },
    {
      "epoch": 1.2251978813307147,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 5147
    },
    {
      "epoch": 1.225435934059394,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 5148
    },
    {
      "epoch": 1.2256739867880735,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 5149
    },
    {
      "epoch": 1.2259120395167529,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3307,
      "step": 5150
    },
    {
      "epoch": 1.2261500922454323,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.595,
      "step": 5151
    },
    {
      "epoch": 1.2263881449741119,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 5152
    },
    {
      "epoch": 1.2266261977027912,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 5153
    },
    {
      "epoch": 1.2268642504314706,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 5154
    },
    {
      "epoch": 1.22710230316015,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2455,
      "step": 5155
    },
    {
      "epoch": 1.2273403558888294,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8575,
      "step": 5156
    },
    {
      "epoch": 1.2275784086175088,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 5157
    },
    {
      "epoch": 1.2278164613461882,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3167,
      "step": 5158
    },
    {
      "epoch": 1.2280545140748675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 5159
    },
    {
      "epoch": 1.228292566803547,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4034,
      "step": 5160
    },
    {
      "epoch": 1.2285306195322263,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 5161
    },
    {
      "epoch": 1.2287686722609057,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3848,
      "step": 5162
    },
    {
      "epoch": 1.2290067249895853,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.5961,
      "step": 5163
    },
    {
      "epoch": 1.2292447777182647,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 5164
    },
    {
      "epoch": 1.229482830446944,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.5072,
      "step": 5165
    },
    {
      "epoch": 1.2297208831756234,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 5166
    },
    {
      "epoch": 1.2299589359043028,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 5167
    },
    {
      "epoch": 1.2301969886329822,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 5168
    },
    {
      "epoch": 1.2304350413616616,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 5169
    },
    {
      "epoch": 1.230673094090341,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 5170
    },
    {
      "epoch": 1.2309111468190204,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3468,
      "step": 5171
    },
    {
      "epoch": 1.2311491995476997,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4263,
      "step": 5172
    },
    {
      "epoch": 1.2313872522763791,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 5173
    },
    {
      "epoch": 1.2316253050050587,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 5174
    },
    {
      "epoch": 1.2318633577337381,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 5175
    },
    {
      "epoch": 1.2321014104624175,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 5176
    },
    {
      "epoch": 1.2323394631910969,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3955,
      "step": 5177
    },
    {
      "epoch": 1.2325775159197763,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 5178
    },
    {
      "epoch": 1.2328155686484557,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5695,
      "step": 5179
    },
    {
      "epoch": 1.233053621377135,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 5180
    },
    {
      "epoch": 1.2332916741058144,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 5181
    },
    {
      "epoch": 1.2335297268344938,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 5182
    },
    {
      "epoch": 1.2337677795631732,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 5183
    },
    {
      "epoch": 1.2340058322918526,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 5184
    },
    {
      "epoch": 1.2342438850205322,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 5185
    },
    {
      "epoch": 1.2344819377492113,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 5186
    },
    {
      "epoch": 1.234719990477891,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 5187
    },
    {
      "epoch": 1.2349580432065703,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 5188
    },
    {
      "epoch": 1.2351960959352497,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 5189
    },
    {
      "epoch": 1.235434148663929,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.393,
      "step": 5190
    },
    {
      "epoch": 1.2356722013926085,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.5902,
      "step": 5191
    },
    {
      "epoch": 1.2359102541212879,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 5192
    },
    {
      "epoch": 1.2361483068499672,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 5193
    },
    {
      "epoch": 1.2363863595786466,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 5194
    },
    {
      "epoch": 1.236624412307326,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2631,
      "step": 5195
    },
    {
      "epoch": 1.2368624650360054,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3491,
      "step": 5196
    },
    {
      "epoch": 1.2371005177646848,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4149,
      "step": 5197
    },
    {
      "epoch": 1.2373385704933644,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 5198
    },
    {
      "epoch": 1.2375766232220438,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 5199
    },
    {
      "epoch": 1.2378146759507231,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 5200
    },
    {
      "epoch": 1.2378146759507231,
      "eval_runtime": 0.5095,
      "eval_samples_per_second": 1.963,
      "eval_steps_per_second": 1.963,
      "step": 5200
    },
    {
      "epoch": 1.2380527286794025,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 5201
    },
    {
      "epoch": 1.238290781408082,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 5202
    },
    {
      "epoch": 1.2385288341367613,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 5203
    },
    {
      "epoch": 1.2387668868654407,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 5204
    },
    {
      "epoch": 1.23900493959412,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 5205
    },
    {
      "epoch": 1.2392429923227994,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 5206
    },
    {
      "epoch": 1.2394810450514788,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 5207
    },
    {
      "epoch": 1.2397190977801582,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 5208
    },
    {
      "epoch": 1.2399571505088378,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4327,
      "step": 5209
    },
    {
      "epoch": 1.2401952032375172,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 5210
    },
    {
      "epoch": 1.2404332559661966,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 5211
    },
    {
      "epoch": 1.240671308694876,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 5212
    },
    {
      "epoch": 1.2409093614235553,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 5213
    },
    {
      "epoch": 1.2411474141522347,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 5214
    },
    {
      "epoch": 1.2413854668809141,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 5215
    },
    {
      "epoch": 1.2416235196095935,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 5216
    },
    {
      "epoch": 1.2418615723382729,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 5217
    },
    {
      "epoch": 1.2420996250669523,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 5218
    },
    {
      "epoch": 1.2423376777956316,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 5219
    },
    {
      "epoch": 1.2425757305243113,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 5220
    },
    {
      "epoch": 1.2428137832529906,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 5221
    },
    {
      "epoch": 1.24305183598167,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 5222
    },
    {
      "epoch": 1.2432898887103494,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 5223
    },
    {
      "epoch": 1.2435279414390288,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 5224
    },
    {
      "epoch": 1.2437659941677082,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 5225
    },
    {
      "epoch": 1.2440040468963875,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 5226
    },
    {
      "epoch": 1.244242099625067,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 5227
    },
    {
      "epoch": 1.2444801523537463,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 5228
    },
    {
      "epoch": 1.2447182050824257,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4077,
      "step": 5229
    },
    {
      "epoch": 1.244956257811105,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3061,
      "step": 5230
    },
    {
      "epoch": 1.2451943105397845,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 5231
    },
    {
      "epoch": 1.2454323632684638,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 5232
    },
    {
      "epoch": 1.2456704159971435,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4174,
      "step": 5233
    },
    {
      "epoch": 1.2459084687258228,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 5234
    },
    {
      "epoch": 1.2461465214545022,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 5235
    },
    {
      "epoch": 1.2463845741831816,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 5236
    },
    {
      "epoch": 1.246622626911861,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 5237
    },
    {
      "epoch": 1.2468606796405404,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 5238
    },
    {
      "epoch": 1.2470987323692198,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 5239
    },
    {
      "epoch": 1.2473367850978991,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 5240
    },
    {
      "epoch": 1.2475748378265785,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 5241
    },
    {
      "epoch": 1.247812890555258,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.399,
      "step": 5242
    },
    {
      "epoch": 1.2480509432839373,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2435,
      "step": 5243
    },
    {
      "epoch": 1.248288996012617,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 5244
    },
    {
      "epoch": 1.2485270487412963,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 5245
    },
    {
      "epoch": 1.2487651014699757,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 5246
    },
    {
      "epoch": 1.249003154198655,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 5247
    },
    {
      "epoch": 1.2492412069273344,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 5248
    },
    {
      "epoch": 1.2494792596560138,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 5249
    },
    {
      "epoch": 1.2497173123846932,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4257,
      "step": 5250
    },
    {
      "epoch": 1.2499553651133726,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3877,
      "step": 5251
    },
    {
      "epoch": 1.250193417842052,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.5057,
      "step": 5252
    },
    {
      "epoch": 1.2504314705707313,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3245,
      "step": 5253
    },
    {
      "epoch": 1.2506695232994107,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 5254
    },
    {
      "epoch": 1.2509075760280903,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.6953,
      "step": 5255
    },
    {
      "epoch": 1.2511456287567695,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 5256
    },
    {
      "epoch": 1.251383681485449,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.4279,
      "step": 5257
    },
    {
      "epoch": 1.2516217342141285,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3726,
      "step": 5258
    },
    {
      "epoch": 1.2518597869428079,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 5259
    },
    {
      "epoch": 1.2520978396714872,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 5260
    },
    {
      "epoch": 1.2523358924001666,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 5261
    },
    {
      "epoch": 1.252573945128846,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 5262
    },
    {
      "epoch": 1.2528119978575254,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 5263
    },
    {
      "epoch": 1.2530500505862048,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 5264
    },
    {
      "epoch": 1.2532881033148842,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 5265
    },
    {
      "epoch": 1.2535261560435638,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 5266
    },
    {
      "epoch": 1.253764208772243,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 5267
    },
    {
      "epoch": 1.2540022615009225,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 5268
    },
    {
      "epoch": 1.254240314229602,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 5269
    },
    {
      "epoch": 1.2544783669582813,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 5270
    },
    {
      "epoch": 1.2547164196869607,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 5271
    },
    {
      "epoch": 1.25495447241564,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 5272
    },
    {
      "epoch": 1.2551925251443194,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 5273
    },
    {
      "epoch": 1.2554305778729988,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 5274
    },
    {
      "epoch": 1.2556686306016782,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2496,
      "step": 5275
    },
    {
      "epoch": 1.2559066833303576,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 5276
    },
    {
      "epoch": 1.2561447360590372,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 5277
    },
    {
      "epoch": 1.2563827887877164,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 5278
    },
    {
      "epoch": 1.256620841516396,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 5279
    },
    {
      "epoch": 1.2568588942450754,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.346,
      "step": 5280
    },
    {
      "epoch": 1.2570969469737547,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.7289,
      "step": 5281
    },
    {
      "epoch": 1.2573349997024341,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 5282
    },
    {
      "epoch": 1.2575730524311135,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.433,
      "step": 5283
    },
    {
      "epoch": 1.2578111051597929,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 5284
    },
    {
      "epoch": 1.2580491578884723,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 5285
    },
    {
      "epoch": 1.2582872106171517,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3238,
      "step": 5286
    },
    {
      "epoch": 1.258525263345831,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 5287
    },
    {
      "epoch": 1.2587633160745106,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 5288
    },
    {
      "epoch": 1.2590013688031898,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3978,
      "step": 5289
    },
    {
      "epoch": 1.2592394215318694,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3202,
      "step": 5290
    },
    {
      "epoch": 1.2594774742605488,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 5291
    },
    {
      "epoch": 1.2597155269892282,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 5292
    },
    {
      "epoch": 1.2599535797179076,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2528,
      "step": 5293
    },
    {
      "epoch": 1.260191632446587,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 5294
    },
    {
      "epoch": 1.2604296851752663,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 5295
    },
    {
      "epoch": 1.2606677379039457,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 5296
    },
    {
      "epoch": 1.260905790632625,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 5297
    },
    {
      "epoch": 1.2611438433613045,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 5298
    },
    {
      "epoch": 1.2613818960899839,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 5299
    },
    {
      "epoch": 1.2616199488186632,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 5300
    },
    {
      "epoch": 1.2616199488186632,
      "eval_runtime": 0.5204,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 5300
    },
    {
      "epoch": 1.2618580015473428,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3134,
      "step": 5301
    },
    {
      "epoch": 1.262096054276022,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 5302
    },
    {
      "epoch": 1.2623341070047016,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 5303
    },
    {
      "epoch": 1.262572159733381,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 5304
    },
    {
      "epoch": 1.2628102124620604,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 5305
    },
    {
      "epoch": 1.2630482651907398,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 5306
    },
    {
      "epoch": 1.2632863179194191,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 5307
    },
    {
      "epoch": 1.2635243706480985,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 5308
    },
    {
      "epoch": 1.263762423376778,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 5309
    },
    {
      "epoch": 1.2640004761054573,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 5310
    },
    {
      "epoch": 1.2642385288341367,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 5311
    },
    {
      "epoch": 1.2644765815628163,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3638,
      "step": 5312
    },
    {
      "epoch": 1.2647146342914954,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1956,
      "step": 5313
    },
    {
      "epoch": 1.264952687020175,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 5314
    },
    {
      "epoch": 1.2651907397488544,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4287,
      "step": 5315
    },
    {
      "epoch": 1.2654287924775338,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 5316
    },
    {
      "epoch": 1.2656668452062132,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 5317
    },
    {
      "epoch": 1.2659048979348926,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2955,
      "step": 5318
    },
    {
      "epoch": 1.266142950663572,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 5319
    },
    {
      "epoch": 1.2663810033922513,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 5320
    },
    {
      "epoch": 1.2666190561209307,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 5321
    },
    {
      "epoch": 1.2668571088496101,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2768,
      "step": 5322
    },
    {
      "epoch": 1.2670951615782897,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3234,
      "step": 5323
    },
    {
      "epoch": 1.2673332143069689,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 5324
    },
    {
      "epoch": 1.2675712670356485,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.411,
      "step": 5325
    },
    {
      "epoch": 1.2678093197643279,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 5326
    },
    {
      "epoch": 1.2680473724930073,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.605,
      "step": 5327
    },
    {
      "epoch": 1.2682854252216866,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 5328
    },
    {
      "epoch": 1.268523477950366,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 5329
    },
    {
      "epoch": 1.2687615306790454,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3655,
      "step": 5330
    },
    {
      "epoch": 1.2689995834077248,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 5331
    },
    {
      "epoch": 1.2692376361364042,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 5332
    },
    {
      "epoch": 1.2694756888650836,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 5333
    },
    {
      "epoch": 1.2697137415937632,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 5334
    },
    {
      "epoch": 1.2699517943224423,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 5335
    },
    {
      "epoch": 1.270189847051122,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 5336
    },
    {
      "epoch": 1.2704278997798013,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3339,
      "step": 5337
    },
    {
      "epoch": 1.2706659525084807,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 5338
    },
    {
      "epoch": 1.27090400523716,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 5339
    },
    {
      "epoch": 1.2711420579658395,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 5340
    },
    {
      "epoch": 1.2713801106945188,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 5341
    },
    {
      "epoch": 1.2716181634231982,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 5342
    },
    {
      "epoch": 1.2718562161518776,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 5343
    },
    {
      "epoch": 1.272094268880557,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 5344
    },
    {
      "epoch": 1.2723323216092364,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 5345
    },
    {
      "epoch": 1.2725703743379158,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3686,
      "step": 5346
    },
    {
      "epoch": 1.2728084270665954,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 5347
    },
    {
      "epoch": 1.2730464797952745,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 5348
    },
    {
      "epoch": 1.2732845325239541,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 5349
    },
    {
      "epoch": 1.2735225852526335,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3893,
      "step": 5350
    },
    {
      "epoch": 1.273760637981313,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 5351
    },
    {
      "epoch": 1.2739986907099923,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 5352
    },
    {
      "epoch": 1.2742367434386717,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 5353
    },
    {
      "epoch": 1.274474796167351,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3017,
      "step": 5354
    },
    {
      "epoch": 1.2747128488960304,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 5355
    },
    {
      "epoch": 1.2749509016247098,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 5356
    },
    {
      "epoch": 1.2751889543533892,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4165,
      "step": 5357
    },
    {
      "epoch": 1.2754270070820688,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3938,
      "step": 5358
    },
    {
      "epoch": 1.275665059810748,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 5359
    },
    {
      "epoch": 1.2759031125394276,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 5360
    },
    {
      "epoch": 1.276141165268107,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 5361
    },
    {
      "epoch": 1.2763792179967863,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 5362
    },
    {
      "epoch": 1.2766172707254657,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 5363
    },
    {
      "epoch": 1.276855323454145,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 5364
    },
    {
      "epoch": 1.2770933761828245,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 5365
    },
    {
      "epoch": 1.2773314289115039,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 5366
    },
    {
      "epoch": 1.2775694816401832,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 5367
    },
    {
      "epoch": 1.2778075343688626,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 5368
    },
    {
      "epoch": 1.2780455870975422,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 5369
    },
    {
      "epoch": 1.2782836398262214,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 5370
    },
    {
      "epoch": 1.278521692554901,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 5371
    },
    {
      "epoch": 1.2787597452835804,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 5372
    },
    {
      "epoch": 1.2789977980122598,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4339,
      "step": 5373
    },
    {
      "epoch": 1.2792358507409392,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 5374
    },
    {
      "epoch": 1.2794739034696185,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 5375
    },
    {
      "epoch": 1.279711956198298,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 5376
    },
    {
      "epoch": 1.2799500089269773,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 5377
    },
    {
      "epoch": 1.2801880616556567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 5378
    },
    {
      "epoch": 1.280426114384336,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 5379
    },
    {
      "epoch": 1.2806641671130155,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 5380
    },
    {
      "epoch": 1.2809022198416948,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 5381
    },
    {
      "epoch": 1.2811402725703744,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3735,
      "step": 5382
    },
    {
      "epoch": 1.2813783252990538,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 5383
    },
    {
      "epoch": 1.2816163780277332,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 5384
    },
    {
      "epoch": 1.2818544307564126,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 5385
    },
    {
      "epoch": 1.282092483485092,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 5386
    },
    {
      "epoch": 1.2823305362137714,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 5387
    },
    {
      "epoch": 1.2825685889424507,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 5388
    },
    {
      "epoch": 1.2828066416711301,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 5389
    },
    {
      "epoch": 1.2830446943998095,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3459,
      "step": 5390
    },
    {
      "epoch": 1.2832827471284889,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 5391
    },
    {
      "epoch": 1.2835207998571683,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 5392
    },
    {
      "epoch": 1.2837588525858479,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 5393
    },
    {
      "epoch": 1.283996905314527,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 5394
    },
    {
      "epoch": 1.2842349580432066,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 5395
    },
    {
      "epoch": 1.284473010771886,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2931,
      "step": 5396
    },
    {
      "epoch": 1.2847110635005654,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 5397
    },
    {
      "epoch": 1.2849491162292448,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 5398
    },
    {
      "epoch": 1.2851871689579242,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 5399
    },
    {
      "epoch": 1.2854252216866036,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 5400
    },
    {
      "epoch": 1.2854252216866036,
      "eval_runtime": 0.5182,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 5400
    },
    {
      "epoch": 1.285663274415283,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 5401
    },
    {
      "epoch": 1.2859013271439623,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 5402
    },
    {
      "epoch": 1.2861393798726417,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 5403
    },
    {
      "epoch": 1.2863774326013213,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 5404
    },
    {
      "epoch": 1.2866154853300005,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.4281,
      "step": 5405
    },
    {
      "epoch": 1.28685353805868,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3183,
      "step": 5406
    },
    {
      "epoch": 1.2870915907873595,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 5407
    },
    {
      "epoch": 1.2873296435160388,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 5408
    },
    {
      "epoch": 1.2875676962447182,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.5908,
      "step": 5409
    },
    {
      "epoch": 1.2878057489733976,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 5410
    },
    {
      "epoch": 1.288043801702077,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 5411
    },
    {
      "epoch": 1.2882818544307564,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 5412
    },
    {
      "epoch": 1.2885199071594358,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3494,
      "step": 5413
    },
    {
      "epoch": 1.2887579598881151,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 5414
    },
    {
      "epoch": 1.2889960126167948,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 5415
    },
    {
      "epoch": 1.289234065345474,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 5416
    },
    {
      "epoch": 1.2894721180741535,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 5417
    },
    {
      "epoch": 1.289710170802833,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 5418
    },
    {
      "epoch": 1.2899482235315123,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 5419
    },
    {
      "epoch": 1.2901862762601917,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3764,
      "step": 5420
    },
    {
      "epoch": 1.290424328988871,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 5421
    },
    {
      "epoch": 1.2906623817175504,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 5422
    },
    {
      "epoch": 1.2909004344462298,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 5423
    },
    {
      "epoch": 1.2911384871749092,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 5424
    },
    {
      "epoch": 1.2913765399035886,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 5425
    },
    {
      "epoch": 1.291614592632268,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 5426
    },
    {
      "epoch": 1.2918526453609473,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 5427
    },
    {
      "epoch": 1.292090698089627,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 5428
    },
    {
      "epoch": 1.2923287508183063,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 5429
    },
    {
      "epoch": 1.2925668035469857,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 5430
    },
    {
      "epoch": 1.292804856275665,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4454,
      "step": 5431
    },
    {
      "epoch": 1.2930429090043445,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5676,
      "step": 5432
    },
    {
      "epoch": 1.2932809617330239,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3199,
      "step": 5433
    },
    {
      "epoch": 1.2935190144617033,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 5434
    },
    {
      "epoch": 1.2937570671903826,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 5435
    },
    {
      "epoch": 1.293995119919062,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5085,
      "step": 5436
    },
    {
      "epoch": 1.2942331726477414,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 5437
    },
    {
      "epoch": 1.2944712253764208,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 5438
    },
    {
      "epoch": 1.2947092781051004,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 5439
    },
    {
      "epoch": 1.2949473308337796,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 5440
    },
    {
      "epoch": 1.2951853835624592,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 5441
    },
    {
      "epoch": 1.2954234362911385,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 5442
    },
    {
      "epoch": 1.295661489019818,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 5443
    },
    {
      "epoch": 1.2958995417484973,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3404,
      "step": 5444
    },
    {
      "epoch": 1.2961375944771767,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 5445
    },
    {
      "epoch": 1.296375647205856,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 5446
    },
    {
      "epoch": 1.2966136999345355,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 5447
    },
    {
      "epoch": 1.2968517526632148,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 5448
    },
    {
      "epoch": 1.2970898053918942,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 5449
    },
    {
      "epoch": 1.2973278581205738,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 5450
    },
    {
      "epoch": 1.297565910849253,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3957,
      "step": 5451
    },
    {
      "epoch": 1.2978039635779326,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 5452
    },
    {
      "epoch": 1.298042016306612,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3062,
      "step": 5453
    },
    {
      "epoch": 1.2982800690352914,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 5454
    },
    {
      "epoch": 1.2985181217639707,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2997,
      "step": 5455
    },
    {
      "epoch": 1.2987561744926501,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 5456
    },
    {
      "epoch": 1.2989942272213295,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 5457
    },
    {
      "epoch": 1.299232279950009,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 5458
    },
    {
      "epoch": 1.2994703326786883,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 5459
    },
    {
      "epoch": 1.2997083854073677,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 5460
    },
    {
      "epoch": 1.2999464381360473,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 5461
    },
    {
      "epoch": 1.3001844908647264,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4536,
      "step": 5462
    },
    {
      "epoch": 1.300422543593406,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 5463
    },
    {
      "epoch": 1.3006605963220854,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 5464
    },
    {
      "epoch": 1.3008986490507648,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 5465
    },
    {
      "epoch": 1.3011367017794442,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 5466
    },
    {
      "epoch": 1.3013747545081236,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 5467
    },
    {
      "epoch": 1.301612807236803,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.4695,
      "step": 5468
    },
    {
      "epoch": 1.3018508599654823,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 5469
    },
    {
      "epoch": 1.3020889126941617,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 5470
    },
    {
      "epoch": 1.302326965422841,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 5471
    },
    {
      "epoch": 1.3025650181515205,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 5472
    },
    {
      "epoch": 1.3028030708801999,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 5473
    },
    {
      "epoch": 1.3030411236088795,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 5474
    },
    {
      "epoch": 1.3032791763375586,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 5475
    },
    {
      "epoch": 1.3035172290662382,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 5476
    },
    {
      "epoch": 1.3037552817949176,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4105,
      "step": 5477
    },
    {
      "epoch": 1.303993334523597,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 5478
    },
    {
      "epoch": 1.3042313872522764,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3761,
      "step": 5479
    },
    {
      "epoch": 1.3044694399809558,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 5480
    },
    {
      "epoch": 1.3047074927096352,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 5481
    },
    {
      "epoch": 1.3049455454383145,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3533,
      "step": 5482
    },
    {
      "epoch": 1.305183598166994,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 5483
    },
    {
      "epoch": 1.3054216508956733,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3173,
      "step": 5484
    },
    {
      "epoch": 1.305659703624353,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 5485
    },
    {
      "epoch": 1.305897756353032,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 5486
    },
    {
      "epoch": 1.3061358090817117,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3944,
      "step": 5487
    },
    {
      "epoch": 1.306373861810391,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 5488
    },
    {
      "epoch": 1.3066119145390704,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4551,
      "step": 5489
    },
    {
      "epoch": 1.3068499672677498,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 5490
    },
    {
      "epoch": 1.3070880199964292,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 5491
    },
    {
      "epoch": 1.3073260727251086,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 5492
    },
    {
      "epoch": 1.307564125453788,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 5493
    },
    {
      "epoch": 1.3078021781824674,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 5494
    },
    {
      "epoch": 1.3080402309111467,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 5495
    },
    {
      "epoch": 1.3082782836398263,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 5496
    },
    {
      "epoch": 1.3085163363685055,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 5497
    },
    {
      "epoch": 1.3087543890971851,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3899,
      "step": 5498
    },
    {
      "epoch": 1.3089924418258645,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3528,
      "step": 5499
    },
    {
      "epoch": 1.3092304945545439,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 5500
    },
    {
      "epoch": 1.3092304945545439,
      "eval_runtime": 0.5104,
      "eval_samples_per_second": 1.959,
      "eval_steps_per_second": 1.959,
      "step": 5500
    },
    {
      "epoch": 1.3094685472832233,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 5501
    },
    {
      "epoch": 1.3097066000119026,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 5502
    },
    {
      "epoch": 1.309944652740582,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 5503
    },
    {
      "epoch": 1.3101827054692614,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3782,
      "step": 5504
    },
    {
      "epoch": 1.3104207581979408,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 5505
    },
    {
      "epoch": 1.3106588109266202,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 5506
    },
    {
      "epoch": 1.3108968636552998,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4619,
      "step": 5507
    },
    {
      "epoch": 1.311134916383979,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 5508
    },
    {
      "epoch": 1.3113729691126585,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 5509
    },
    {
      "epoch": 1.311611021841338,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 5510
    },
    {
      "epoch": 1.3118490745700173,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 5511
    },
    {
      "epoch": 1.3120871272986967,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 5512
    },
    {
      "epoch": 1.312325180027376,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 5513
    },
    {
      "epoch": 1.3125632327560555,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 5514
    },
    {
      "epoch": 1.3128012854847348,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 5515
    },
    {
      "epoch": 1.3130393382134142,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 5516
    },
    {
      "epoch": 1.3132773909420936,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 5517
    },
    {
      "epoch": 1.313515443670773,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 5518
    },
    {
      "epoch": 1.3137534963994524,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 5519
    },
    {
      "epoch": 1.313991549128132,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 5520
    },
    {
      "epoch": 1.3142296018568111,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 5521
    },
    {
      "epoch": 1.3144676545854908,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3361,
      "step": 5522
    },
    {
      "epoch": 1.3147057073141701,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 5523
    },
    {
      "epoch": 1.3149437600428495,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 5524
    },
    {
      "epoch": 1.315181812771529,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 5525
    },
    {
      "epoch": 1.3154198655002083,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 5526
    },
    {
      "epoch": 1.3156579182288877,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 5527
    },
    {
      "epoch": 1.315895970957567,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 5528
    },
    {
      "epoch": 1.3161340236862464,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5298,
      "step": 5529
    },
    {
      "epoch": 1.3163720764149258,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 5530
    },
    {
      "epoch": 1.3166101291436054,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 5531
    },
    {
      "epoch": 1.3168481818722846,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 5532
    },
    {
      "epoch": 1.3170862346009642,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 5533
    },
    {
      "epoch": 1.3173242873296436,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2549,
      "step": 5534
    },
    {
      "epoch": 1.317562340058323,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 5535
    },
    {
      "epoch": 1.3178003927870023,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 5536
    },
    {
      "epoch": 1.3180384455156817,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 5537
    },
    {
      "epoch": 1.318276498244361,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 5538
    },
    {
      "epoch": 1.3185145509730405,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 5539
    },
    {
      "epoch": 1.3187526037017199,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8488,
      "step": 5540
    },
    {
      "epoch": 1.3189906564303993,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 5541
    },
    {
      "epoch": 1.3192287091590789,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 5542
    },
    {
      "epoch": 1.319466761887758,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 5543
    },
    {
      "epoch": 1.3197048146164376,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.751,
      "step": 5544
    },
    {
      "epoch": 1.319942867345117,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 5545
    },
    {
      "epoch": 1.3201809200737964,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3802,
      "step": 5546
    },
    {
      "epoch": 1.3204189728024758,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 5547
    },
    {
      "epoch": 1.3206570255311552,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2605,
      "step": 5548
    },
    {
      "epoch": 1.3208950782598345,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3908,
      "step": 5549
    },
    {
      "epoch": 1.321133130988514,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 5550
    },
    {
      "epoch": 1.3213711837171933,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 5551
    },
    {
      "epoch": 1.3216092364458727,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 5552
    },
    {
      "epoch": 1.3218472891745523,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 5553
    },
    {
      "epoch": 1.3220853419032315,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 5554
    },
    {
      "epoch": 1.322323394631911,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.3548,
      "step": 5555
    },
    {
      "epoch": 1.3225614473605904,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3469,
      "step": 5556
    },
    {
      "epoch": 1.3227995000892698,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 5557
    },
    {
      "epoch": 1.3230375528179492,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 5558
    },
    {
      "epoch": 1.3232756055466286,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 5559
    },
    {
      "epoch": 1.323513658275308,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 5560
    },
    {
      "epoch": 1.3237517110039874,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 5561
    },
    {
      "epoch": 1.3239897637326667,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3724,
      "step": 5562
    },
    {
      "epoch": 1.3242278164613461,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3591,
      "step": 5563
    },
    {
      "epoch": 1.3244658691900255,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 5564
    },
    {
      "epoch": 1.324703921918705,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 5565
    },
    {
      "epoch": 1.3249419746473845,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 5566
    },
    {
      "epoch": 1.3251800273760637,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 5567
    },
    {
      "epoch": 1.3254180801047433,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3338,
      "step": 5568
    },
    {
      "epoch": 1.3256561328334227,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2507,
      "step": 5569
    },
    {
      "epoch": 1.325894185562102,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 5570
    },
    {
      "epoch": 1.3261322382907814,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3791,
      "step": 5571
    },
    {
      "epoch": 1.3263702910194608,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 5572
    },
    {
      "epoch": 1.3266083437481402,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3606,
      "step": 5573
    },
    {
      "epoch": 1.3268463964768196,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 5574
    },
    {
      "epoch": 1.327084449205499,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 5575
    },
    {
      "epoch": 1.3273225019341783,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 5576
    },
    {
      "epoch": 1.327560554662858,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4153,
      "step": 5577
    },
    {
      "epoch": 1.327798607391537,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 5578
    },
    {
      "epoch": 1.3280366601202167,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3996,
      "step": 5579
    },
    {
      "epoch": 1.328274712848896,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 5580
    },
    {
      "epoch": 1.3285127655775755,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 5581
    },
    {
      "epoch": 1.3287508183062549,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 5582
    },
    {
      "epoch": 1.3289888710349342,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 5583
    },
    {
      "epoch": 1.3292269237636136,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 5584
    },
    {
      "epoch": 1.329464976492293,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 5585
    },
    {
      "epoch": 1.3297030292209724,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 5586
    },
    {
      "epoch": 1.3299410819496518,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 5587
    },
    {
      "epoch": 1.3301791346783314,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 5588
    },
    {
      "epoch": 1.3304171874070105,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 5589
    },
    {
      "epoch": 1.3306552401356901,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 5590
    },
    {
      "epoch": 1.3308932928643695,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2544,
      "step": 5591
    },
    {
      "epoch": 1.331131345593049,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4357,
      "step": 5592
    },
    {
      "epoch": 1.3313693983217283,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 5593
    },
    {
      "epoch": 1.3316074510504077,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 5594
    },
    {
      "epoch": 1.331845503779087,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 5595
    },
    {
      "epoch": 1.3320835565077664,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 5596
    },
    {
      "epoch": 1.3323216092364458,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 5597
    },
    {
      "epoch": 1.3325596619651252,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 5598
    },
    {
      "epoch": 1.3327977146938048,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 5599
    },
    {
      "epoch": 1.333035767422484,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 5600
    },
    {
      "epoch": 1.333035767422484,
      "eval_runtime": 0.5139,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 5600
    },
    {
      "epoch": 1.3332738201511636,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 5601
    },
    {
      "epoch": 1.333511872879843,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 5602
    },
    {
      "epoch": 1.3337499256085223,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 5603
    },
    {
      "epoch": 1.3339879783372017,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 5604
    },
    {
      "epoch": 1.3342260310658811,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 5605
    },
    {
      "epoch": 1.3344640837945605,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 5606
    },
    {
      "epoch": 1.3347021365232399,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 5607
    },
    {
      "epoch": 1.3349401892519193,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 5608
    },
    {
      "epoch": 1.3351782419805986,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 5609
    },
    {
      "epoch": 1.335416294709278,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 5610
    },
    {
      "epoch": 1.3356543474379574,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5635,
      "step": 5611
    },
    {
      "epoch": 1.335892400166637,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 5612
    },
    {
      "epoch": 1.3361304528953162,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3534,
      "step": 5613
    },
    {
      "epoch": 1.3363685056239958,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 5614
    },
    {
      "epoch": 1.3366065583526752,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2555,
      "step": 5615
    },
    {
      "epoch": 1.3368446110813546,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 5616
    },
    {
      "epoch": 1.337082663810034,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.366,
      "step": 5617
    },
    {
      "epoch": 1.3373207165387133,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3407,
      "step": 5618
    },
    {
      "epoch": 1.3375587692673927,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 5619
    },
    {
      "epoch": 1.337796821996072,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3694,
      "step": 5620
    },
    {
      "epoch": 1.3380348747247515,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 5621
    },
    {
      "epoch": 1.3382729274534308,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 5622
    },
    {
      "epoch": 1.3385109801821105,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4184,
      "step": 5623
    },
    {
      "epoch": 1.3387490329107896,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 5624
    },
    {
      "epoch": 1.3389870856394692,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4063,
      "step": 5625
    },
    {
      "epoch": 1.3392251383681486,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 5626
    },
    {
      "epoch": 1.339463191096828,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 5627
    },
    {
      "epoch": 1.3397012438255074,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 5628
    },
    {
      "epoch": 1.3399392965541868,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 5629
    },
    {
      "epoch": 1.3401773492828661,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 5630
    },
    {
      "epoch": 1.3404154020115455,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4884,
      "step": 5631
    },
    {
      "epoch": 1.340653454740225,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 5632
    },
    {
      "epoch": 1.3408915074689043,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3517,
      "step": 5633
    },
    {
      "epoch": 1.341129560197584,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 5634
    },
    {
      "epoch": 1.341367612926263,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 5635
    },
    {
      "epoch": 1.3416056656549427,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 5636
    },
    {
      "epoch": 1.341843718383622,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3792,
      "step": 5637
    },
    {
      "epoch": 1.3420817711123014,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 5638
    },
    {
      "epoch": 1.3423198238409808,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.476,
      "step": 5639
    },
    {
      "epoch": 1.3425578765696602,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 5640
    },
    {
      "epoch": 1.3427959292983396,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 5641
    },
    {
      "epoch": 1.343033982027019,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 5642
    },
    {
      "epoch": 1.3432720347556983,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 5643
    },
    {
      "epoch": 1.3435100874843777,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 5644
    },
    {
      "epoch": 1.3437481402130573,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3044,
      "step": 5645
    },
    {
      "epoch": 1.3439861929417365,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 5646
    },
    {
      "epoch": 1.344224245670416,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 5647
    },
    {
      "epoch": 1.3444622983990955,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 5648
    },
    {
      "epoch": 1.3447003511277749,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 5649
    },
    {
      "epoch": 1.3449384038564542,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 5650
    },
    {
      "epoch": 1.3451764565851336,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4056,
      "step": 5651
    },
    {
      "epoch": 1.345414509313813,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 5652
    },
    {
      "epoch": 1.3456525620424924,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 5653
    },
    {
      "epoch": 1.3458906147711718,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 5654
    },
    {
      "epoch": 1.3461286674998512,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 5655
    },
    {
      "epoch": 1.3463667202285305,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 5656
    },
    {
      "epoch": 1.34660477295721,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 5657
    },
    {
      "epoch": 1.3468428256858895,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3253,
      "step": 5658
    },
    {
      "epoch": 1.3470808784145687,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 5659
    },
    {
      "epoch": 1.3473189311432483,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 5660
    },
    {
      "epoch": 1.3475569838719277,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 5661
    },
    {
      "epoch": 1.347795036600607,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 5662
    },
    {
      "epoch": 1.3480330893292864,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 5663
    },
    {
      "epoch": 1.3482711420579658,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 5664
    },
    {
      "epoch": 1.3485091947866452,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 5665
    },
    {
      "epoch": 1.3487472475153246,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 5666
    },
    {
      "epoch": 1.348985300244004,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 5667
    },
    {
      "epoch": 1.3492233529726834,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 5668
    },
    {
      "epoch": 1.349461405701363,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 5669
    },
    {
      "epoch": 1.3496994584300421,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 5670
    },
    {
      "epoch": 1.3499375111587217,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3234,
      "step": 5671
    },
    {
      "epoch": 1.3501755638874011,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 5672
    },
    {
      "epoch": 1.3504136166160805,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 5673
    },
    {
      "epoch": 1.3506516693447599,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 5674
    },
    {
      "epoch": 1.3508897220734393,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 5675
    },
    {
      "epoch": 1.3511277748021187,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 5676
    },
    {
      "epoch": 1.351365827530798,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 5677
    },
    {
      "epoch": 1.3516038802594774,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 5678
    },
    {
      "epoch": 1.3518419329881568,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 5679
    },
    {
      "epoch": 1.3520799857168364,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 5680
    },
    {
      "epoch": 1.3523180384455156,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 5681
    },
    {
      "epoch": 1.3525560911741952,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4106,
      "step": 5682
    },
    {
      "epoch": 1.3527941439028746,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 5683
    },
    {
      "epoch": 1.353032196631554,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 5684
    },
    {
      "epoch": 1.3532702493602333,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 5685
    },
    {
      "epoch": 1.3535083020889127,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 5686
    },
    {
      "epoch": 1.353746354817592,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 5687
    },
    {
      "epoch": 1.3539844075462715,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4484,
      "step": 5688
    },
    {
      "epoch": 1.3542224602749509,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 5689
    },
    {
      "epoch": 1.3544605130036302,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 5690
    },
    {
      "epoch": 1.3546985657323098,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 5691
    },
    {
      "epoch": 1.354936618460989,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 5692
    },
    {
      "epoch": 1.3551746711896686,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4686,
      "step": 5693
    },
    {
      "epoch": 1.355412723918348,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3907,
      "step": 5694
    },
    {
      "epoch": 1.3556507766470274,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3285,
      "step": 5695
    },
    {
      "epoch": 1.3558888293757068,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 5696
    },
    {
      "epoch": 1.3561268821043861,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 5697
    },
    {
      "epoch": 1.3563649348330655,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 5698
    },
    {
      "epoch": 1.356602987561745,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3495,
      "step": 5699
    },
    {
      "epoch": 1.3568410402904243,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 5700
    },
    {
      "epoch": 1.3568410402904243,
      "eval_runtime": 0.5683,
      "eval_samples_per_second": 1.76,
      "eval_steps_per_second": 1.76,
      "step": 5700
    },
    {
      "epoch": 1.3570790930191037,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 5701
    },
    {
      "epoch": 1.357317145747783,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 5702
    },
    {
      "epoch": 1.3575551984764624,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 5703
    },
    {
      "epoch": 1.357793251205142,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3741,
      "step": 5704
    },
    {
      "epoch": 1.3580313039338212,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 5705
    },
    {
      "epoch": 1.3582693566625008,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 5706
    },
    {
      "epoch": 1.3585074093911802,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 5707
    },
    {
      "epoch": 1.3587454621198596,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3483,
      "step": 5708
    },
    {
      "epoch": 1.358983514848539,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 5709
    },
    {
      "epoch": 1.3592215675772183,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 5710
    },
    {
      "epoch": 1.3594596203058977,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 5711
    },
    {
      "epoch": 1.3596976730345771,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 5712
    },
    {
      "epoch": 1.3599357257632565,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 5713
    },
    {
      "epoch": 1.3601737784919359,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 5714
    },
    {
      "epoch": 1.3604118312206155,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 5715
    },
    {
      "epoch": 1.3606498839492946,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3505,
      "step": 5716
    },
    {
      "epoch": 1.3608879366779743,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 5717
    },
    {
      "epoch": 1.3611259894066536,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 5718
    },
    {
      "epoch": 1.361364042135333,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 5719
    },
    {
      "epoch": 1.3616020948640124,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 5720
    },
    {
      "epoch": 1.3618401475926918,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4259,
      "step": 5721
    },
    {
      "epoch": 1.3620782003213712,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 5722
    },
    {
      "epoch": 1.3623162530500506,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 5723
    },
    {
      "epoch": 1.36255430577873,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2438,
      "step": 5724
    },
    {
      "epoch": 1.3627923585074093,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 5725
    },
    {
      "epoch": 1.363030411236089,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 5726
    },
    {
      "epoch": 1.363268463964768,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 5727
    },
    {
      "epoch": 1.3635065166934477,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 5728
    },
    {
      "epoch": 1.363744569422127,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3856,
      "step": 5729
    },
    {
      "epoch": 1.3639826221508065,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 5730
    },
    {
      "epoch": 1.3642206748794858,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 5731
    },
    {
      "epoch": 1.3644587276081652,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 5732
    },
    {
      "epoch": 1.3646967803368446,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 5733
    },
    {
      "epoch": 1.364934833065524,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 5734
    },
    {
      "epoch": 1.3651728857942034,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 5735
    },
    {
      "epoch": 1.3654109385228828,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 5736
    },
    {
      "epoch": 1.3656489912515621,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 5737
    },
    {
      "epoch": 1.3658870439802415,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 5738
    },
    {
      "epoch": 1.3661250967089211,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3222,
      "step": 5739
    },
    {
      "epoch": 1.3663631494376005,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 5740
    },
    {
      "epoch": 1.36660120216628,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 5741
    },
    {
      "epoch": 1.3668392548949593,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2791,
      "step": 5742
    },
    {
      "epoch": 1.3670773076236387,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 5743
    },
    {
      "epoch": 1.367315360352318,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 5744
    },
    {
      "epoch": 1.3675534130809974,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 5745
    },
    {
      "epoch": 1.3677914658096768,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 5746
    },
    {
      "epoch": 1.3680295185383562,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 5747
    },
    {
      "epoch": 1.3682675712670356,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 5748
    },
    {
      "epoch": 1.368505623995715,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 5749
    },
    {
      "epoch": 1.3687436767243946,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 5750
    },
    {
      "epoch": 1.3689817294530737,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 5751
    },
    {
      "epoch": 1.3692197821817533,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 5752
    },
    {
      "epoch": 1.3694578349104327,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 5753
    },
    {
      "epoch": 1.369695887639112,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 5754
    },
    {
      "epoch": 1.3699339403677915,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4016,
      "step": 5755
    },
    {
      "epoch": 1.3701719930964709,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 5756
    },
    {
      "epoch": 1.3704100458251502,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 5757
    },
    {
      "epoch": 1.3706480985538296,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 5758
    },
    {
      "epoch": 1.370886151282509,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 5759
    },
    {
      "epoch": 1.3711242040111884,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2591,
      "step": 5760
    },
    {
      "epoch": 1.371362256739868,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 5761
    },
    {
      "epoch": 1.3716003094685472,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 5762
    },
    {
      "epoch": 1.3718383621972268,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 5763
    },
    {
      "epoch": 1.3720764149259062,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 5764
    },
    {
      "epoch": 1.3723144676545855,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 5765
    },
    {
      "epoch": 1.372552520383265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 5766
    },
    {
      "epoch": 1.3727905731119443,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 5767
    },
    {
      "epoch": 1.3730286258406237,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 5768
    },
    {
      "epoch": 1.373266678569303,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 5769
    },
    {
      "epoch": 1.3735047312979825,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 5770
    },
    {
      "epoch": 1.3737427840266618,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 5771
    },
    {
      "epoch": 1.3739808367553414,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4702,
      "step": 5772
    },
    {
      "epoch": 1.3742188894840206,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 5773
    },
    {
      "epoch": 1.3744569422127002,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 5774
    },
    {
      "epoch": 1.3746949949413796,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 5775
    },
    {
      "epoch": 1.374933047670059,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 5776
    },
    {
      "epoch": 1.3751711003987384,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 5777
    },
    {
      "epoch": 1.3754091531274177,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 5778
    },
    {
      "epoch": 1.3756472058560971,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 5779
    },
    {
      "epoch": 1.3758852585847765,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 5780
    },
    {
      "epoch": 1.3761233113134559,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 5781
    },
    {
      "epoch": 1.3763613640421353,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 5782
    },
    {
      "epoch": 1.3765994167708147,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5162,
      "step": 5783
    },
    {
      "epoch": 1.376837469499494,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.489,
      "step": 5784
    },
    {
      "epoch": 1.3770755222281736,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 5785
    },
    {
      "epoch": 1.377313574956853,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 5786
    },
    {
      "epoch": 1.3775516276855324,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 5787
    },
    {
      "epoch": 1.3777896804142118,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 5788
    },
    {
      "epoch": 1.3780277331428912,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 5789
    },
    {
      "epoch": 1.3782657858715706,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 5790
    },
    {
      "epoch": 1.37850383860025,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 5791
    },
    {
      "epoch": 1.3787418913289293,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4441,
      "step": 5792
    },
    {
      "epoch": 1.3789799440576087,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 5793
    },
    {
      "epoch": 1.379217996786288,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4738,
      "step": 5794
    },
    {
      "epoch": 1.3794560495149675,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 5795
    },
    {
      "epoch": 1.379694102243647,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 5796
    },
    {
      "epoch": 1.3799321549723262,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3616,
      "step": 5797
    },
    {
      "epoch": 1.3801702077010058,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 5798
    },
    {
      "epoch": 1.3804082604296852,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 5799
    },
    {
      "epoch": 1.3806463131583646,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 5800
    },
    {
      "epoch": 1.3806463131583646,
      "eval_runtime": 0.5343,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 1.872,
      "step": 5800
    },
    {
      "epoch": 1.380884365887044,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3371,
      "step": 5801
    },
    {
      "epoch": 1.3811224186157234,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 5802
    },
    {
      "epoch": 1.3813604713444028,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 5803
    },
    {
      "epoch": 1.3815985240730821,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 5804
    },
    {
      "epoch": 1.3818365768017615,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 5805
    },
    {
      "epoch": 1.382074629530441,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 5806
    },
    {
      "epoch": 1.3823126822591205,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 5807
    },
    {
      "epoch": 1.3825507349877997,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 5808
    },
    {
      "epoch": 1.3827887877164793,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 5809
    },
    {
      "epoch": 1.3830268404451587,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 5810
    },
    {
      "epoch": 1.383264893173838,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 5811
    },
    {
      "epoch": 1.3835029459025174,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 5812
    },
    {
      "epoch": 1.3837409986311968,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 5813
    },
    {
      "epoch": 1.3839790513598762,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 5814
    },
    {
      "epoch": 1.3842171040885556,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 5815
    },
    {
      "epoch": 1.384455156817235,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 5816
    },
    {
      "epoch": 1.3846932095459143,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 5817
    },
    {
      "epoch": 1.384931262274594,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 5818
    },
    {
      "epoch": 1.3851693150032731,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 5819
    },
    {
      "epoch": 1.3854073677319527,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 5820
    },
    {
      "epoch": 1.385645420460632,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 5821
    },
    {
      "epoch": 1.3858834731893115,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 5822
    },
    {
      "epoch": 1.3861215259179909,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3538,
      "step": 5823
    },
    {
      "epoch": 1.3863595786466703,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 5824
    },
    {
      "epoch": 1.3865976313753496,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 5825
    },
    {
      "epoch": 1.386835684104029,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 5826
    },
    {
      "epoch": 1.3870737368327084,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 5827
    },
    {
      "epoch": 1.3873117895613878,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.4984,
      "step": 5828
    },
    {
      "epoch": 1.3875498422900672,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2526,
      "step": 5829
    },
    {
      "epoch": 1.3877878950187466,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 5830
    },
    {
      "epoch": 1.3880259477474262,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4128,
      "step": 5831
    },
    {
      "epoch": 1.3882640004761053,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 5832
    },
    {
      "epoch": 1.388502053204785,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 5833
    },
    {
      "epoch": 1.3887401059334643,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 5834
    },
    {
      "epoch": 1.3889781586621437,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3762,
      "step": 5835
    },
    {
      "epoch": 1.389216211390823,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.505,
      "step": 5836
    },
    {
      "epoch": 1.3894542641195025,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 5837
    },
    {
      "epoch": 1.3896923168481818,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.6969,
      "step": 5838
    },
    {
      "epoch": 1.3899303695768612,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 5839
    },
    {
      "epoch": 1.3901684223055406,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 5840
    },
    {
      "epoch": 1.39040647503422,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 5841
    },
    {
      "epoch": 1.3906445277628996,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 5842
    },
    {
      "epoch": 1.3908825804915788,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 5843
    },
    {
      "epoch": 1.3911206332202584,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 5844
    },
    {
      "epoch": 1.3913586859489377,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 5845
    },
    {
      "epoch": 1.3915967386776171,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 5846
    },
    {
      "epoch": 1.3918347914062965,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 5847
    },
    {
      "epoch": 1.392072844134976,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 5848
    },
    {
      "epoch": 1.3923108968636553,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 5849
    },
    {
      "epoch": 1.3925489495923347,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 5850
    },
    {
      "epoch": 1.392787002321014,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 5851
    },
    {
      "epoch": 1.3930250550496934,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 5852
    },
    {
      "epoch": 1.393263107778373,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.6079,
      "step": 5853
    },
    {
      "epoch": 1.3935011605070522,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 5854
    },
    {
      "epoch": 1.3937392132357318,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 5855
    },
    {
      "epoch": 1.3939772659644112,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 5856
    },
    {
      "epoch": 1.3942153186930906,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 5857
    },
    {
      "epoch": 1.39445337142177,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2437,
      "step": 5858
    },
    {
      "epoch": 1.3946914241504493,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3363,
      "step": 5859
    },
    {
      "epoch": 1.3949294768791287,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 5860
    },
    {
      "epoch": 1.395167529607808,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 5861
    },
    {
      "epoch": 1.3954055823364875,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 5862
    },
    {
      "epoch": 1.3956436350651669,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 5863
    },
    {
      "epoch": 1.3958816877938465,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.286,
      "step": 5864
    },
    {
      "epoch": 1.3961197405225256,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 5865
    },
    {
      "epoch": 1.3963577932512052,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 5866
    },
    {
      "epoch": 1.3965958459798846,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 5867
    },
    {
      "epoch": 1.396833898708564,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 5868
    },
    {
      "epoch": 1.3970719514372434,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.4613,
      "step": 5869
    },
    {
      "epoch": 1.3973100041659228,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 5870
    },
    {
      "epoch": 1.3975480568946022,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 5871
    },
    {
      "epoch": 1.3977861096232815,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 5872
    },
    {
      "epoch": 1.398024162351961,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 5873
    },
    {
      "epoch": 1.3982622150806403,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 5874
    },
    {
      "epoch": 1.3985002678093197,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 5875
    },
    {
      "epoch": 1.398738320537999,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 5876
    },
    {
      "epoch": 1.3989763732666787,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 5877
    },
    {
      "epoch": 1.3992144259953578,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 5878
    },
    {
      "epoch": 1.3994524787240374,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 5879
    },
    {
      "epoch": 1.3996905314527168,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 5880
    },
    {
      "epoch": 1.3999285841813962,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 5881
    },
    {
      "epoch": 1.4001666369100756,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 5882
    },
    {
      "epoch": 1.400404689638755,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 5883
    },
    {
      "epoch": 1.4006427423674344,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 5884
    },
    {
      "epoch": 1.4008807950961137,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4645,
      "step": 5885
    },
    {
      "epoch": 1.4011188478247931,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.276,
      "step": 5886
    },
    {
      "epoch": 1.4013569005534725,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 5887
    },
    {
      "epoch": 1.4015949532821521,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 5888
    },
    {
      "epoch": 1.4018330060108313,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 5889
    },
    {
      "epoch": 1.4020710587395109,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 5890
    },
    {
      "epoch": 1.4023091114681903,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 5891
    },
    {
      "epoch": 1.4025471641968696,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 5892
    },
    {
      "epoch": 1.402785216925549,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 5893
    },
    {
      "epoch": 1.4030232696542284,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3212,
      "step": 5894
    },
    {
      "epoch": 1.4032613223829078,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.5039,
      "step": 5895
    },
    {
      "epoch": 1.4034993751115872,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 5896
    },
    {
      "epoch": 1.4037374278402666,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 5897
    },
    {
      "epoch": 1.403975480568946,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3529,
      "step": 5898
    },
    {
      "epoch": 1.4042135332976255,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 5899
    },
    {
      "epoch": 1.4044515860263047,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 5900
    },
    {
      "epoch": 1.4044515860263047,
      "eval_runtime": 0.5051,
      "eval_samples_per_second": 1.98,
      "eval_steps_per_second": 1.98,
      "step": 5900
    },
    {
      "epoch": 1.4046896387549843,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 5901
    },
    {
      "epoch": 1.4049276914836637,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 5902
    },
    {
      "epoch": 1.405165744212343,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 5903
    },
    {
      "epoch": 1.4054037969410225,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 5904
    },
    {
      "epoch": 1.4056418496697018,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 5905
    },
    {
      "epoch": 1.4058799023983812,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 5906
    },
    {
      "epoch": 1.4061179551270606,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 5907
    },
    {
      "epoch": 1.40635600785574,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 5908
    },
    {
      "epoch": 1.4065940605844194,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3917,
      "step": 5909
    },
    {
      "epoch": 1.406832113313099,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 5910
    },
    {
      "epoch": 1.4070701660417781,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4333,
      "step": 5911
    },
    {
      "epoch": 1.4073082187704578,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 5912
    },
    {
      "epoch": 1.4075462714991371,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 5913
    },
    {
      "epoch": 1.4077843242278165,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 5914
    },
    {
      "epoch": 1.408022376956496,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 5915
    },
    {
      "epoch": 1.4082604296851753,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 5916
    },
    {
      "epoch": 1.4084984824138547,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 5917
    },
    {
      "epoch": 1.408736535142534,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 5918
    },
    {
      "epoch": 1.4089745878712134,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.6326,
      "step": 5919
    },
    {
      "epoch": 1.4092126405998928,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 5920
    },
    {
      "epoch": 1.4094506933285722,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 5921
    },
    {
      "epoch": 1.4096887460572516,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 5922
    },
    {
      "epoch": 1.4099267987859312,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 5923
    },
    {
      "epoch": 1.4101648515146104,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 5924
    },
    {
      "epoch": 1.41040290424329,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3714,
      "step": 5925
    },
    {
      "epoch": 1.4106409569719693,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 5926
    },
    {
      "epoch": 1.4108790097006487,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 5927
    },
    {
      "epoch": 1.411117062429328,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 5928
    },
    {
      "epoch": 1.4113551151580075,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3977,
      "step": 5929
    },
    {
      "epoch": 1.4115931678866869,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 5930
    },
    {
      "epoch": 1.4118312206153663,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.6539,
      "step": 5931
    },
    {
      "epoch": 1.4120692733440456,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 5932
    },
    {
      "epoch": 1.412307326072725,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 5933
    },
    {
      "epoch": 1.4125453788014046,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.6438,
      "step": 5934
    },
    {
      "epoch": 1.4127834315300838,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 5935
    },
    {
      "epoch": 1.4130214842587634,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 5936
    },
    {
      "epoch": 1.4132595369874428,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4238,
      "step": 5937
    },
    {
      "epoch": 1.4134975897161222,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4189,
      "step": 5938
    },
    {
      "epoch": 1.4137356424448015,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 5939
    },
    {
      "epoch": 1.413973695173481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 5940
    },
    {
      "epoch": 1.4142117479021603,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 5941
    },
    {
      "epoch": 1.4144498006308397,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 5942
    },
    {
      "epoch": 1.414687853359519,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 5943
    },
    {
      "epoch": 1.4149259060881985,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2744,
      "step": 5944
    },
    {
      "epoch": 1.415163958816878,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.6121,
      "step": 5945
    },
    {
      "epoch": 1.4154020115455572,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4539,
      "step": 5946
    },
    {
      "epoch": 1.4156400642742368,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 5947
    },
    {
      "epoch": 1.4158781170029162,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 5948
    },
    {
      "epoch": 1.4161161697315956,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3339,
      "step": 5949
    },
    {
      "epoch": 1.416354222460275,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 5950
    },
    {
      "epoch": 1.4165922751889544,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 5951
    },
    {
      "epoch": 1.4168303279176337,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 5952
    },
    {
      "epoch": 1.4170683806463131,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 5953
    },
    {
      "epoch": 1.4173064333749925,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.307,
      "step": 5954
    },
    {
      "epoch": 1.417544486103672,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 5955
    },
    {
      "epoch": 1.4177825388323515,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 5956
    },
    {
      "epoch": 1.4180205915610307,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 5957
    },
    {
      "epoch": 1.4182586442897103,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 5958
    },
    {
      "epoch": 1.4184966970183897,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 5959
    },
    {
      "epoch": 1.418734749747069,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 5960
    },
    {
      "epoch": 1.4189728024757484,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3926,
      "step": 5961
    },
    {
      "epoch": 1.4192108552044278,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3205,
      "step": 5962
    },
    {
      "epoch": 1.4194489079331072,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 5963
    },
    {
      "epoch": 1.4196869606617866,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 5964
    },
    {
      "epoch": 1.419925013390466,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 5965
    },
    {
      "epoch": 1.4201630661191453,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 5966
    },
    {
      "epoch": 1.4204011188478247,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 5967
    },
    {
      "epoch": 1.420639171576504,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 5968
    },
    {
      "epoch": 1.4208772243051837,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 5969
    },
    {
      "epoch": 1.4211152770338629,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 5970
    },
    {
      "epoch": 1.4213533297625425,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 5971
    },
    {
      "epoch": 1.4215913824912219,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3493,
      "step": 5972
    },
    {
      "epoch": 1.4218294352199012,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3843,
      "step": 5973
    },
    {
      "epoch": 1.4220674879485806,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 5974
    },
    {
      "epoch": 1.42230554067726,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4665,
      "step": 5975
    },
    {
      "epoch": 1.4225435934059394,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.5394,
      "step": 5976
    },
    {
      "epoch": 1.4227816461346188,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 5977
    },
    {
      "epoch": 1.4230196988632982,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 5978
    },
    {
      "epoch": 1.4232577515919775,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 5979
    },
    {
      "epoch": 1.4234958043206571,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 5980
    },
    {
      "epoch": 1.4237338570493363,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 5981
    },
    {
      "epoch": 1.423971909778016,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 5982
    },
    {
      "epoch": 1.4242099625066953,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 5983
    },
    {
      "epoch": 1.4244480152353747,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3916,
      "step": 5984
    },
    {
      "epoch": 1.424686067964054,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 5985
    },
    {
      "epoch": 1.4249241206927334,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4263,
      "step": 5986
    },
    {
      "epoch": 1.4251621734214128,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 5987
    },
    {
      "epoch": 1.4254002261500922,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 5988
    },
    {
      "epoch": 1.4256382788787716,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 5989
    },
    {
      "epoch": 1.425876331607451,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 5990
    },
    {
      "epoch": 1.4261143843361306,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 5991
    },
    {
      "epoch": 1.4263524370648097,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2647,
      "step": 5992
    },
    {
      "epoch": 1.4265904897934893,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 5993
    },
    {
      "epoch": 1.4268285425221687,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.3943,
      "step": 5994
    },
    {
      "epoch": 1.4270665952508481,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4825,
      "step": 5995
    },
    {
      "epoch": 1.4273046479795275,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 5996
    },
    {
      "epoch": 1.4275427007082069,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 5997
    },
    {
      "epoch": 1.4277807534368863,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 5998
    },
    {
      "epoch": 1.4280188061655656,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 5999
    },
    {
      "epoch": 1.428256858894245,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 6000
    },
    {
      "epoch": 1.428256858894245,
      "eval_runtime": 0.5152,
      "eval_samples_per_second": 1.941,
      "eval_steps_per_second": 1.941,
      "step": 6000
    },
    {
      "epoch": 1.4284949116229244,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 6001
    },
    {
      "epoch": 1.428732964351604,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 6002
    },
    {
      "epoch": 1.4289710170802832,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3818,
      "step": 6003
    },
    {
      "epoch": 1.4292090698089628,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 6004
    },
    {
      "epoch": 1.4294471225376422,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 6005
    },
    {
      "epoch": 1.4296851752663216,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 6006
    },
    {
      "epoch": 1.429923227995001,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.4806,
      "step": 6007
    },
    {
      "epoch": 1.4301612807236803,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.4451,
      "step": 6008
    },
    {
      "epoch": 1.4303993334523597,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 6009
    },
    {
      "epoch": 1.430637386181039,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 6010
    },
    {
      "epoch": 1.4308754389097185,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 6011
    },
    {
      "epoch": 1.4311134916383978,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 6012
    },
    {
      "epoch": 1.4313515443670772,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 6013
    },
    {
      "epoch": 1.4315895970957566,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2365,
      "step": 6014
    },
    {
      "epoch": 1.4318276498244362,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3158,
      "step": 6015
    },
    {
      "epoch": 1.4320657025531154,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 6016
    },
    {
      "epoch": 1.432303755281795,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3526,
      "step": 6017
    },
    {
      "epoch": 1.4325418080104744,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 6018
    },
    {
      "epoch": 1.4327798607391538,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 6019
    },
    {
      "epoch": 1.4330179134678331,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 6020
    },
    {
      "epoch": 1.4332559661965125,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 6021
    },
    {
      "epoch": 1.433494018925192,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 6022
    },
    {
      "epoch": 1.4337320716538713,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5319,
      "step": 6023
    },
    {
      "epoch": 1.4339701243825507,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 6024
    },
    {
      "epoch": 1.43420817711123,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3418,
      "step": 6025
    },
    {
      "epoch": 1.4344462298399097,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 6026
    },
    {
      "epoch": 1.4346842825685888,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 6027
    },
    {
      "epoch": 1.4349223352972684,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 6028
    },
    {
      "epoch": 1.4351603880259478,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.357,
      "step": 6029
    },
    {
      "epoch": 1.4353984407546272,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 6030
    },
    {
      "epoch": 1.4356364934833066,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 6031
    },
    {
      "epoch": 1.435874546211986,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4285,
      "step": 6032
    },
    {
      "epoch": 1.4361125989406653,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 6033
    },
    {
      "epoch": 1.4363506516693447,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 6034
    },
    {
      "epoch": 1.436588704398024,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 6035
    },
    {
      "epoch": 1.4368267571267035,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3064,
      "step": 6036
    },
    {
      "epoch": 1.437064809855383,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 6037
    },
    {
      "epoch": 1.4373028625840623,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 6038
    },
    {
      "epoch": 1.4375409153127419,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 6039
    },
    {
      "epoch": 1.4377789680414212,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 6040
    },
    {
      "epoch": 1.4380170207701006,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 6041
    },
    {
      "epoch": 1.43825507349878,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 6042
    },
    {
      "epoch": 1.4384931262274594,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 6043
    },
    {
      "epoch": 1.4387311789561388,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 6044
    },
    {
      "epoch": 1.4389692316848182,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 6045
    },
    {
      "epoch": 1.4392072844134975,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 6046
    },
    {
      "epoch": 1.439445337142177,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 6047
    },
    {
      "epoch": 1.4396833898708563,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 6048
    },
    {
      "epoch": 1.4399214425995357,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3442,
      "step": 6049
    },
    {
      "epoch": 1.4401594953282153,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 6050
    },
    {
      "epoch": 1.4403975480568947,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.4127,
      "step": 6051
    },
    {
      "epoch": 1.440635600785574,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 6052
    },
    {
      "epoch": 1.4408736535142534,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 6053
    },
    {
      "epoch": 1.4411117062429328,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4179,
      "step": 6054
    },
    {
      "epoch": 1.4413497589716122,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 6055
    },
    {
      "epoch": 1.4415878117002916,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 6056
    },
    {
      "epoch": 1.441825864428971,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 6057
    },
    {
      "epoch": 1.4420639171576504,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 6058
    },
    {
      "epoch": 1.4423019698863297,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 6059
    },
    {
      "epoch": 1.4425400226150091,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 6060
    },
    {
      "epoch": 1.4427780753436887,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 6061
    },
    {
      "epoch": 1.443016128072368,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 6062
    },
    {
      "epoch": 1.4432541808010475,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 6063
    },
    {
      "epoch": 1.4434922335297269,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3845,
      "step": 6064
    },
    {
      "epoch": 1.4437302862584063,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2828,
      "step": 6065
    },
    {
      "epoch": 1.4439683389870857,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 6066
    },
    {
      "epoch": 1.444206391715765,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 6067
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 6068
    },
    {
      "epoch": 1.4446824971731238,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 6069
    },
    {
      "epoch": 1.4449205499018032,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3831,
      "step": 6070
    },
    {
      "epoch": 1.4451586026304826,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 6071
    },
    {
      "epoch": 1.4453966553591622,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 6072
    },
    {
      "epoch": 1.4456347080878413,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 6073
    },
    {
      "epoch": 1.445872760816521,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 6074
    },
    {
      "epoch": 1.4461108135452003,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4016,
      "step": 6075
    },
    {
      "epoch": 1.4463488662738797,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 6076
    },
    {
      "epoch": 1.446586919002559,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 6077
    },
    {
      "epoch": 1.4468249717312385,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 6078
    },
    {
      "epoch": 1.4470630244599179,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3657,
      "step": 6079
    },
    {
      "epoch": 1.4473010771885972,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 6080
    },
    {
      "epoch": 1.4475391299172766,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 6081
    },
    {
      "epoch": 1.447777182645956,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 6082
    },
    {
      "epoch": 1.4480152353746356,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.7913,
      "step": 6083
    },
    {
      "epoch": 1.4482532881033148,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 6084
    },
    {
      "epoch": 1.4484913408319944,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 6085
    },
    {
      "epoch": 1.4487293935606738,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 6086
    },
    {
      "epoch": 1.4489674462893531,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 6087
    },
    {
      "epoch": 1.4492054990180325,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 6088
    },
    {
      "epoch": 1.449443551746712,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 6089
    },
    {
      "epoch": 1.4496816044753913,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.4286,
      "step": 6090
    },
    {
      "epoch": 1.4499196572040707,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 6091
    },
    {
      "epoch": 1.45015770993275,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3161,
      "step": 6092
    },
    {
      "epoch": 1.4503957626614294,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4652,
      "step": 6093
    },
    {
      "epoch": 1.4506338153901088,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 6094
    },
    {
      "epoch": 1.4508718681187882,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 6095
    },
    {
      "epoch": 1.4511099208474678,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 6096
    },
    {
      "epoch": 1.4513479735761472,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.313,
      "step": 6097
    },
    {
      "epoch": 1.4515860263048266,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 6098
    },
    {
      "epoch": 1.451824079033506,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4765,
      "step": 6099
    },
    {
      "epoch": 1.4520621317621853,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 6100
    },
    {
      "epoch": 1.4520621317621853,
      "eval_runtime": 0.4996,
      "eval_samples_per_second": 2.001,
      "eval_steps_per_second": 2.001,
      "step": 6100
    },
    {
      "epoch": 1.4523001844908647,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3854,
      "step": 6101
    },
    {
      "epoch": 1.4525382372195441,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 6102
    },
    {
      "epoch": 1.4527762899482235,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2206,
      "step": 6103
    },
    {
      "epoch": 1.4530143426769029,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 6104
    },
    {
      "epoch": 1.4532523954055823,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 6105
    },
    {
      "epoch": 1.4534904481342616,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 6106
    },
    {
      "epoch": 1.4537285008629413,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 6107
    },
    {
      "epoch": 1.4539665535916204,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 6108
    },
    {
      "epoch": 1.4542046063203,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4664,
      "step": 6109
    },
    {
      "epoch": 1.4544426590489794,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 6110
    },
    {
      "epoch": 1.4546807117776588,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 6111
    },
    {
      "epoch": 1.4549187645063382,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 6112
    },
    {
      "epoch": 1.4551568172350176,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 6113
    },
    {
      "epoch": 1.455394869963697,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 6114
    },
    {
      "epoch": 1.4556329226923763,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 6115
    },
    {
      "epoch": 1.4558709754210557,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 6116
    },
    {
      "epoch": 1.456109028149735,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 6117
    },
    {
      "epoch": 1.4563470808784147,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 6118
    },
    {
      "epoch": 1.4565851336070939,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 6119
    },
    {
      "epoch": 1.4568231863357735,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 6120
    },
    {
      "epoch": 1.4570612390644528,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 6121
    },
    {
      "epoch": 1.4572992917931322,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 6122
    },
    {
      "epoch": 1.4575373445218116,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 6123
    },
    {
      "epoch": 1.457775397250491,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 6124
    },
    {
      "epoch": 1.4580134499791704,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3183,
      "step": 6125
    },
    {
      "epoch": 1.4582515027078498,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 6126
    },
    {
      "epoch": 1.4584895554365291,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 6127
    },
    {
      "epoch": 1.4587276081652085,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 6128
    },
    {
      "epoch": 1.4589656608938881,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 6129
    },
    {
      "epoch": 1.4592037136225673,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 6130
    },
    {
      "epoch": 1.459441766351247,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 6131
    },
    {
      "epoch": 1.4596798190799263,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5422,
      "step": 6132
    },
    {
      "epoch": 1.4599178718086057,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 6133
    },
    {
      "epoch": 1.460155924537285,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 6134
    },
    {
      "epoch": 1.4603939772659644,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 6135
    },
    {
      "epoch": 1.4606320299946438,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 6136
    },
    {
      "epoch": 1.4608700827233232,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 6137
    },
    {
      "epoch": 1.4611081354520026,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 6138
    },
    {
      "epoch": 1.461346188180682,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3753,
      "step": 6139
    },
    {
      "epoch": 1.4615842409093613,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 6140
    },
    {
      "epoch": 1.4618222936380407,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3673,
      "step": 6141
    },
    {
      "epoch": 1.4620603463667203,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 6142
    },
    {
      "epoch": 1.4622983990953995,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.355,
      "step": 6143
    },
    {
      "epoch": 1.462536451824079,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 6144
    },
    {
      "epoch": 1.4627745045527585,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 6145
    },
    {
      "epoch": 1.4630125572814379,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 6146
    },
    {
      "epoch": 1.4632506100101172,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 6147
    },
    {
      "epoch": 1.4634886627387966,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.593,
      "step": 6148
    },
    {
      "epoch": 1.463726715467476,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 6149
    },
    {
      "epoch": 1.4639647681961554,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 6150
    },
    {
      "epoch": 1.4642028209248348,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 6151
    },
    {
      "epoch": 1.4644408736535142,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 6152
    },
    {
      "epoch": 1.4646789263821938,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.5985,
      "step": 6153
    },
    {
      "epoch": 1.464916979110873,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 6154
    },
    {
      "epoch": 1.4651550318395525,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3544,
      "step": 6155
    },
    {
      "epoch": 1.465393084568232,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 6156
    },
    {
      "epoch": 1.4656311372969113,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 6157
    },
    {
      "epoch": 1.4658691900255907,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2627,
      "step": 6158
    },
    {
      "epoch": 1.46610724275427,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4242,
      "step": 6159
    },
    {
      "epoch": 1.4663452954829495,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 6160
    },
    {
      "epoch": 1.4665833482116288,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 6161
    },
    {
      "epoch": 1.4668214009403082,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 6162
    },
    {
      "epoch": 1.4670594536689876,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 6163
    },
    {
      "epoch": 1.4672975063976672,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 6164
    },
    {
      "epoch": 1.4675355591263464,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 6165
    },
    {
      "epoch": 1.467773611855026,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 6166
    },
    {
      "epoch": 1.4680116645837054,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 6167
    },
    {
      "epoch": 1.4682497173123847,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4585,
      "step": 6168
    },
    {
      "epoch": 1.4684877700410641,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2038,
      "step": 6169
    },
    {
      "epoch": 1.4687258227697435,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 6170
    },
    {
      "epoch": 1.4689638754984229,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 6171
    },
    {
      "epoch": 1.4692019282271023,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 6172
    },
    {
      "epoch": 1.4694399809557817,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 6173
    },
    {
      "epoch": 1.469678033684461,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 6174
    },
    {
      "epoch": 1.4699160864131406,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3679,
      "step": 6175
    },
    {
      "epoch": 1.4701541391418198,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 6176
    },
    {
      "epoch": 1.4703921918704994,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 6177
    },
    {
      "epoch": 1.4706302445991788,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.4404,
      "step": 6178
    },
    {
      "epoch": 1.4708682973278582,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2812,
      "step": 6179
    },
    {
      "epoch": 1.4711063500565376,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 6180
    },
    {
      "epoch": 1.471344402785217,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 6181
    },
    {
      "epoch": 1.4715824555138963,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 6182
    },
    {
      "epoch": 1.4718205082425757,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 6183
    },
    {
      "epoch": 1.472058560971255,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 6184
    },
    {
      "epoch": 1.4722966136999345,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4054,
      "step": 6185
    },
    {
      "epoch": 1.4725346664286139,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 6186
    },
    {
      "epoch": 1.4727727191572932,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.5398,
      "step": 6187
    },
    {
      "epoch": 1.4730107718859728,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 6188
    },
    {
      "epoch": 1.473248824614652,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 6189
    },
    {
      "epoch": 1.4734868773433316,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 6190
    },
    {
      "epoch": 1.473724930072011,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 6191
    },
    {
      "epoch": 1.4739629828006904,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 6192
    },
    {
      "epoch": 1.4742010355293698,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4068,
      "step": 6193
    },
    {
      "epoch": 1.4744390882580491,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 6194
    },
    {
      "epoch": 1.4746771409867285,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 6195
    },
    {
      "epoch": 1.474915193715408,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 6196
    },
    {
      "epoch": 1.4751532464440873,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 6197
    },
    {
      "epoch": 1.4753912991727667,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 6198
    },
    {
      "epoch": 1.4756293519014463,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 6199
    },
    {
      "epoch": 1.4758674046301254,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 6200
    },
    {
      "epoch": 1.4758674046301254,
      "eval_runtime": 0.5081,
      "eval_samples_per_second": 1.968,
      "eval_steps_per_second": 1.968,
      "step": 6200
    },
    {
      "epoch": 1.476105457358805,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 6201
    },
    {
      "epoch": 1.4763435100874844,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 6202
    },
    {
      "epoch": 1.4765815628161638,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 6203
    },
    {
      "epoch": 1.4768196155448432,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 6204
    },
    {
      "epoch": 1.4770576682735226,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 6205
    },
    {
      "epoch": 1.477295721002202,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 6206
    },
    {
      "epoch": 1.4775337737308814,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 6207
    },
    {
      "epoch": 1.4777718264595607,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 6208
    },
    {
      "epoch": 1.4780098791882401,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2785,
      "step": 6209
    },
    {
      "epoch": 1.4782479319169197,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 6210
    },
    {
      "epoch": 1.4784859846455989,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 6211
    },
    {
      "epoch": 1.4787240373742785,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 6212
    },
    {
      "epoch": 1.4789620901029579,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4928,
      "step": 6213
    },
    {
      "epoch": 1.4792001428316373,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 6214
    },
    {
      "epoch": 1.4794381955603166,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 6215
    },
    {
      "epoch": 1.479676248288996,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.273,
      "step": 6216
    },
    {
      "epoch": 1.4799143010176754,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 6217
    },
    {
      "epoch": 1.4801523537463548,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 6218
    },
    {
      "epoch": 1.4803904064750342,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3682,
      "step": 6219
    },
    {
      "epoch": 1.4806284592037136,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 6220
    },
    {
      "epoch": 1.4808665119323932,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 6221
    },
    {
      "epoch": 1.4811045646610723,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 6222
    },
    {
      "epoch": 1.481342617389752,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 6223
    },
    {
      "epoch": 1.4815806701184313,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 6224
    },
    {
      "epoch": 1.4818187228471107,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4857,
      "step": 6225
    },
    {
      "epoch": 1.48205677557579,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 6226
    },
    {
      "epoch": 1.4822948283044695,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 6227
    },
    {
      "epoch": 1.4825328810331488,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 6228
    },
    {
      "epoch": 1.4827709337618282,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 6229
    },
    {
      "epoch": 1.4830089864905076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 6230
    },
    {
      "epoch": 1.483247039219187,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3399,
      "step": 6231
    },
    {
      "epoch": 1.4834850919478664,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 6232
    },
    {
      "epoch": 1.4837231446765458,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 6233
    },
    {
      "epoch": 1.4839611974052254,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 6234
    },
    {
      "epoch": 1.4841992501339045,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 6235
    },
    {
      "epoch": 1.4844373028625841,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 6236
    },
    {
      "epoch": 1.4846753555912635,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 6237
    },
    {
      "epoch": 1.484913408319943,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 6238
    },
    {
      "epoch": 1.4851514610486223,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 6239
    },
    {
      "epoch": 1.4853895137773017,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 6240
    },
    {
      "epoch": 1.485627566505981,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 6241
    },
    {
      "epoch": 1.4858656192346604,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 6242
    },
    {
      "epoch": 1.4861036719633398,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4419,
      "step": 6243
    },
    {
      "epoch": 1.4863417246920192,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 6244
    },
    {
      "epoch": 1.4865797774206988,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 6245
    },
    {
      "epoch": 1.486817830149378,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 6246
    },
    {
      "epoch": 1.4870558828780576,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 6247
    },
    {
      "epoch": 1.487293935606737,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 6248
    },
    {
      "epoch": 1.4875319883354163,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 6249
    },
    {
      "epoch": 1.4877700410640957,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3859,
      "step": 6250
    },
    {
      "epoch": 1.488008093792775,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 6251
    },
    {
      "epoch": 1.4882461465214545,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 6252
    },
    {
      "epoch": 1.4884841992501339,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 6253
    },
    {
      "epoch": 1.4887222519788132,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 6254
    },
    {
      "epoch": 1.4889603047074926,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 6255
    },
    {
      "epoch": 1.4891983574361722,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3805,
      "step": 6256
    },
    {
      "epoch": 1.4894364101648514,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 6257
    },
    {
      "epoch": 1.489674462893531,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 6258
    },
    {
      "epoch": 1.4899125156222104,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 6259
    },
    {
      "epoch": 1.4901505683508898,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4413,
      "step": 6260
    },
    {
      "epoch": 1.4903886210795692,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4196,
      "step": 6261
    },
    {
      "epoch": 1.4906266738082485,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 6262
    },
    {
      "epoch": 1.490864726536928,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 6263
    },
    {
      "epoch": 1.4911027792656073,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 6264
    },
    {
      "epoch": 1.4913408319942867,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.5343,
      "step": 6265
    },
    {
      "epoch": 1.491578884722966,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 6266
    },
    {
      "epoch": 1.4918169374516457,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 6267
    },
    {
      "epoch": 1.4920549901803248,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 6268
    },
    {
      "epoch": 1.4922930429090044,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 6269
    },
    {
      "epoch": 1.4925310956376838,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 6270
    },
    {
      "epoch": 1.4927691483663632,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3419,
      "step": 6271
    },
    {
      "epoch": 1.4930072010950426,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 6272
    },
    {
      "epoch": 1.493245253823722,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 6273
    },
    {
      "epoch": 1.4934833065524014,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 6274
    },
    {
      "epoch": 1.4937213592810807,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 6275
    },
    {
      "epoch": 1.4939594120097601,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4328,
      "step": 6276
    },
    {
      "epoch": 1.4941974647384395,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3922,
      "step": 6277
    },
    {
      "epoch": 1.494435517467119,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 6278
    },
    {
      "epoch": 1.4946735701957983,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2986,
      "step": 6279
    },
    {
      "epoch": 1.4949116229244779,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 6280
    },
    {
      "epoch": 1.495149675653157,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3885,
      "step": 6281
    },
    {
      "epoch": 1.4953877283818366,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 6282
    },
    {
      "epoch": 1.495625781110516,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 6283
    },
    {
      "epoch": 1.4958638338391954,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 6284
    },
    {
      "epoch": 1.4961018865678748,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 6285
    },
    {
      "epoch": 1.4963399392965542,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 6286
    },
    {
      "epoch": 1.4965779920252336,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4112,
      "step": 6287
    },
    {
      "epoch": 1.496816044753913,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 6288
    },
    {
      "epoch": 1.4970540974825923,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 6289
    },
    {
      "epoch": 1.4972921502112717,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 6290
    },
    {
      "epoch": 1.4975302029399513,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4933,
      "step": 6291
    },
    {
      "epoch": 1.4977682556686305,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 6292
    },
    {
      "epoch": 1.49800630839731,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 6293
    },
    {
      "epoch": 1.4982443611259895,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2822,
      "step": 6294
    },
    {
      "epoch": 1.4984824138546688,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 6295
    },
    {
      "epoch": 1.4987204665833482,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 6296
    },
    {
      "epoch": 1.4989585193120276,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 6297
    },
    {
      "epoch": 1.499196572040707,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 6298
    },
    {
      "epoch": 1.4994346247693864,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 6299
    },
    {
      "epoch": 1.4996726774980658,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 6300
    },
    {
      "epoch": 1.4996726774980658,
      "eval_runtime": 0.5126,
      "eval_samples_per_second": 1.951,
      "eval_steps_per_second": 1.951,
      "step": 6300
    },
    {
      "epoch": 1.4999107302267451,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 6301
    },
    {
      "epoch": 1.5001487829554248,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 6302
    },
    {
      "epoch": 1.500386835684104,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 6303
    },
    {
      "epoch": 1.5006248884127835,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 6304
    },
    {
      "epoch": 1.5008629411414627,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2432,
      "step": 6305
    },
    {
      "epoch": 1.5011009938701423,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 6306
    },
    {
      "epoch": 1.5013390465988217,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3919,
      "step": 6307
    },
    {
      "epoch": 1.501577099327501,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 6308
    },
    {
      "epoch": 1.5018151520561804,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 6309
    },
    {
      "epoch": 1.5020532047848598,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2855,
      "step": 6310
    },
    {
      "epoch": 1.5022912575135392,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 6311
    },
    {
      "epoch": 1.5025293102422186,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 6312
    },
    {
      "epoch": 1.5027673629708982,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 6313
    },
    {
      "epoch": 1.5030054156995774,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 6314
    },
    {
      "epoch": 1.503243468428257,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 6315
    },
    {
      "epoch": 1.5034815211569361,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 6316
    },
    {
      "epoch": 1.5037195738856157,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 6317
    },
    {
      "epoch": 1.503957626614295,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 6318
    },
    {
      "epoch": 1.5041956793429745,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.4038,
      "step": 6319
    },
    {
      "epoch": 1.5044337320716539,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 6320
    },
    {
      "epoch": 1.5046717848003333,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 6321
    },
    {
      "epoch": 1.5049098375290126,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 6322
    },
    {
      "epoch": 1.505147890257692,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.6814,
      "step": 6323
    },
    {
      "epoch": 1.5053859429863716,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 6324
    },
    {
      "epoch": 1.5056239957150508,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 6325
    },
    {
      "epoch": 1.5058620484437304,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.3868,
      "step": 6326
    },
    {
      "epoch": 1.5061001011724096,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 6327
    },
    {
      "epoch": 1.5063381539010892,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 6328
    },
    {
      "epoch": 1.5065762066297685,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 6329
    },
    {
      "epoch": 1.506814259358448,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 6330
    },
    {
      "epoch": 1.5070523120871273,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 6331
    },
    {
      "epoch": 1.5072903648158067,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 6332
    },
    {
      "epoch": 1.507528417544486,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 6333
    },
    {
      "epoch": 1.5077664702731655,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4764,
      "step": 6334
    },
    {
      "epoch": 1.508004523001845,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 6335
    },
    {
      "epoch": 1.5082425757305242,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 6336
    },
    {
      "epoch": 1.5084806284592038,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 6337
    },
    {
      "epoch": 1.508718681187883,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 6338
    },
    {
      "epoch": 1.5089567339165626,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2819,
      "step": 6339
    },
    {
      "epoch": 1.5091947866452418,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 6340
    },
    {
      "epoch": 1.5094328393739214,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 6341
    },
    {
      "epoch": 1.5096708921026007,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 6342
    },
    {
      "epoch": 1.5099089448312801,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2747,
      "step": 6343
    },
    {
      "epoch": 1.5101469975599595,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3923,
      "step": 6344
    },
    {
      "epoch": 1.510385050288639,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 6345
    },
    {
      "epoch": 1.5106231030173183,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 6346
    },
    {
      "epoch": 1.5108611557459977,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 6347
    },
    {
      "epoch": 1.5110992084746773,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3961,
      "step": 6348
    },
    {
      "epoch": 1.5113372612033564,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 6349
    },
    {
      "epoch": 1.511575313932036,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 6350
    },
    {
      "epoch": 1.5118133666607152,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 6351
    },
    {
      "epoch": 1.5120514193893948,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 6352
    },
    {
      "epoch": 1.5122894721180742,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 6353
    },
    {
      "epoch": 1.5125275248467536,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 6354
    },
    {
      "epoch": 1.512765577575433,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2896,
      "step": 6355
    },
    {
      "epoch": 1.5130036303041123,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 6356
    },
    {
      "epoch": 1.5132416830327917,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 6357
    },
    {
      "epoch": 1.513479735761471,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4183,
      "step": 6358
    },
    {
      "epoch": 1.5137177884901507,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2697,
      "step": 6359
    },
    {
      "epoch": 1.5139558412188299,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4201,
      "step": 6360
    },
    {
      "epoch": 1.5141938939475095,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 6361
    },
    {
      "epoch": 1.5144319466761886,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 6362
    },
    {
      "epoch": 1.5146699994048682,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 6363
    },
    {
      "epoch": 1.5149080521335476,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 6364
    },
    {
      "epoch": 1.515146104862227,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 6365
    },
    {
      "epoch": 1.5153841575909064,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 6366
    },
    {
      "epoch": 1.5156222103195858,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 6367
    },
    {
      "epoch": 1.5158602630482652,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 6368
    },
    {
      "epoch": 1.5160983157769445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 6369
    },
    {
      "epoch": 1.5163363685056241,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 6370
    },
    {
      "epoch": 1.5165744212343033,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 6371
    },
    {
      "epoch": 1.516812473962983,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 6372
    },
    {
      "epoch": 1.517050526691662,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 6373
    },
    {
      "epoch": 1.5172885794203417,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 6374
    },
    {
      "epoch": 1.517526632149021,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 6375
    },
    {
      "epoch": 1.5177646848777004,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 6376
    },
    {
      "epoch": 1.5180027376063798,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3942,
      "step": 6377
    },
    {
      "epoch": 1.5182407903350592,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 6378
    },
    {
      "epoch": 1.5184788430637386,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 6379
    },
    {
      "epoch": 1.518716895792418,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 6380
    },
    {
      "epoch": 1.5189549485210976,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 6381
    },
    {
      "epoch": 1.5191930012497767,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 6382
    },
    {
      "epoch": 1.5194310539784563,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4008,
      "step": 6383
    },
    {
      "epoch": 1.5196691067071355,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 6384
    },
    {
      "epoch": 1.5199071594358151,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 6385
    },
    {
      "epoch": 1.5201452121644943,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 6386
    },
    {
      "epoch": 1.5203832648931739,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 6387
    },
    {
      "epoch": 1.5206213176218533,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4136,
      "step": 6388
    },
    {
      "epoch": 1.5208593703505326,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 6389
    },
    {
      "epoch": 1.521097423079212,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 6390
    },
    {
      "epoch": 1.5213354758078914,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3429,
      "step": 6391
    },
    {
      "epoch": 1.5215735285365708,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 6392
    },
    {
      "epoch": 1.5218115812652502,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 6393
    },
    {
      "epoch": 1.5220496339939298,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 6394
    },
    {
      "epoch": 1.522287686722609,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 6395
    },
    {
      "epoch": 1.5225257394512886,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 6396
    },
    {
      "epoch": 1.5227637921799677,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 6397
    },
    {
      "epoch": 1.5230018449086473,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3209,
      "step": 6398
    },
    {
      "epoch": 1.5232398976373267,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3537,
      "step": 6399
    },
    {
      "epoch": 1.523477950366006,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 6400
    },
    {
      "epoch": 1.523477950366006,
      "eval_runtime": 0.5227,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 6400
    },
    {
      "epoch": 1.5237160030946855,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 6401
    },
    {
      "epoch": 1.5239540558233649,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 6402
    },
    {
      "epoch": 1.5241921085520442,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 6403
    },
    {
      "epoch": 1.5244301612807236,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 6404
    },
    {
      "epoch": 1.5246682140094032,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 6405
    },
    {
      "epoch": 1.5249062667380824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 6406
    },
    {
      "epoch": 1.525144319466762,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 6407
    },
    {
      "epoch": 1.5253823721954411,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2565,
      "step": 6408
    },
    {
      "epoch": 1.5256204249241208,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 6409
    },
    {
      "epoch": 1.5258584776528001,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 6410
    },
    {
      "epoch": 1.5260965303814795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 6411
    },
    {
      "epoch": 1.526334583110159,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 6412
    },
    {
      "epoch": 1.5265726358388383,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 6413
    },
    {
      "epoch": 1.5268106885675177,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 6414
    },
    {
      "epoch": 1.527048741296197,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2555,
      "step": 6415
    },
    {
      "epoch": 1.5272867940248767,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 6416
    },
    {
      "epoch": 1.5275248467535558,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 6417
    },
    {
      "epoch": 1.5277628994822354,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 6418
    },
    {
      "epoch": 1.5280009522109146,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 6419
    },
    {
      "epoch": 1.5282390049395942,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 6420
    },
    {
      "epoch": 1.5284770576682736,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3505,
      "step": 6421
    },
    {
      "epoch": 1.528715110396953,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 6422
    },
    {
      "epoch": 1.5289531631256323,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3966,
      "step": 6423
    },
    {
      "epoch": 1.5291912158543117,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 6424
    },
    {
      "epoch": 1.529429268582991,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 6425
    },
    {
      "epoch": 1.5296673213116705,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.3402,
      "step": 6426
    },
    {
      "epoch": 1.52990537404035,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 6427
    },
    {
      "epoch": 1.5301434267690293,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4188,
      "step": 6428
    },
    {
      "epoch": 1.5303814794977089,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4232,
      "step": 6429
    },
    {
      "epoch": 1.530619532226388,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 6430
    },
    {
      "epoch": 1.5308575849550676,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 6431
    },
    {
      "epoch": 1.5310956376837468,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3476,
      "step": 6432
    },
    {
      "epoch": 1.5313336904124264,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 6433
    },
    {
      "epoch": 1.5315717431411058,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 6434
    },
    {
      "epoch": 1.5318097958697852,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 6435
    },
    {
      "epoch": 1.5320478485984645,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 6436
    },
    {
      "epoch": 1.532285901327144,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 6437
    },
    {
      "epoch": 1.5325239540558233,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.447,
      "step": 6438
    },
    {
      "epoch": 1.5327620067845027,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 6439
    },
    {
      "epoch": 1.5330000595131823,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 6440
    },
    {
      "epoch": 1.5332381122418615,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3748,
      "step": 6441
    },
    {
      "epoch": 1.533476164970541,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2628,
      "step": 6442
    },
    {
      "epoch": 1.5337142176992202,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 6443
    },
    {
      "epoch": 1.5339522704278998,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.5208,
      "step": 6444
    },
    {
      "epoch": 1.5341903231565792,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 6445
    },
    {
      "epoch": 1.5344283758852586,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 6446
    },
    {
      "epoch": 1.534666428613938,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 6447
    },
    {
      "epoch": 1.5349044813426174,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 6448
    },
    {
      "epoch": 1.5351425340712967,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 6449
    },
    {
      "epoch": 1.5353805867999761,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 6450
    },
    {
      "epoch": 1.5356186395286557,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 6451
    },
    {
      "epoch": 1.535856692257335,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2939,
      "step": 6452
    },
    {
      "epoch": 1.5360947449860145,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 6453
    },
    {
      "epoch": 1.5363327977146937,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 6454
    },
    {
      "epoch": 1.5365708504433733,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 6455
    },
    {
      "epoch": 1.5368089031720527,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 6456
    },
    {
      "epoch": 1.537046955900732,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 6457
    },
    {
      "epoch": 1.5372850086294114,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 6458
    },
    {
      "epoch": 1.5375230613580908,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 6459
    },
    {
      "epoch": 1.5377611140867702,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 6460
    },
    {
      "epoch": 1.5379991668154496,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5373,
      "step": 6461
    },
    {
      "epoch": 1.5382372195441292,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 6462
    },
    {
      "epoch": 1.5384752722728083,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 6463
    },
    {
      "epoch": 1.538713325001488,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 6464
    },
    {
      "epoch": 1.538951377730167,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3029,
      "step": 6465
    },
    {
      "epoch": 1.5391894304588467,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 6466
    },
    {
      "epoch": 1.539427483187526,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4002,
      "step": 6467
    },
    {
      "epoch": 1.5396655359162055,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 6468
    },
    {
      "epoch": 1.5399035886448849,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 6469
    },
    {
      "epoch": 1.5401416413735642,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 6470
    },
    {
      "epoch": 1.5403796941022436,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 6471
    },
    {
      "epoch": 1.540617746830923,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.6229,
      "step": 6472
    },
    {
      "epoch": 1.5408557995596026,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.458,
      "step": 6473
    },
    {
      "epoch": 1.5410938522882818,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 6474
    },
    {
      "epoch": 1.5413319050169614,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4014,
      "step": 6475
    },
    {
      "epoch": 1.5415699577456405,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 6476
    },
    {
      "epoch": 1.5418080104743201,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 6477
    },
    {
      "epoch": 1.5420460632029993,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3469,
      "step": 6478
    },
    {
      "epoch": 1.542284115931679,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 6479
    },
    {
      "epoch": 1.5425221686603583,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 6480
    },
    {
      "epoch": 1.5427602213890377,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 6481
    },
    {
      "epoch": 1.542998274117717,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 6482
    },
    {
      "epoch": 1.5432363268463964,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.5056,
      "step": 6483
    },
    {
      "epoch": 1.5434743795750758,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 6484
    },
    {
      "epoch": 1.5437124323037552,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 6485
    },
    {
      "epoch": 1.5439504850324348,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 6486
    },
    {
      "epoch": 1.544188537761114,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 6487
    },
    {
      "epoch": 1.5444265904897936,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4014,
      "step": 6488
    },
    {
      "epoch": 1.5446646432184727,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 6489
    },
    {
      "epoch": 1.5449026959471523,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 6490
    },
    {
      "epoch": 1.5451407486758317,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 6491
    },
    {
      "epoch": 1.5453788014045111,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 6492
    },
    {
      "epoch": 1.5456168541331905,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 6493
    },
    {
      "epoch": 1.5458549068618699,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 6494
    },
    {
      "epoch": 1.5460929595905493,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 6495
    },
    {
      "epoch": 1.5463310123192286,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 6496
    },
    {
      "epoch": 1.5465690650479083,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 6497
    },
    {
      "epoch": 1.5468071177765874,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 6498
    },
    {
      "epoch": 1.547045170505267,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 6499
    },
    {
      "epoch": 1.5472832232339462,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 6500
    },
    {
      "epoch": 1.5472832232339462,
      "eval_runtime": 0.5144,
      "eval_samples_per_second": 1.944,
      "eval_steps_per_second": 1.944,
      "step": 6500
    },
    {
      "epoch": 1.5475212759626258,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 6501
    },
    {
      "epoch": 1.5477593286913052,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 6502
    },
    {
      "epoch": 1.5479973814199846,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 6503
    },
    {
      "epoch": 1.548235434148664,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 6504
    },
    {
      "epoch": 1.5484734868773433,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 6505
    },
    {
      "epoch": 1.5487115396060227,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 6506
    },
    {
      "epoch": 1.548949592334702,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 6507
    },
    {
      "epoch": 1.5491876450633817,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 6508
    },
    {
      "epoch": 1.5494256977920609,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 6509
    },
    {
      "epoch": 1.5496637505207405,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 6510
    },
    {
      "epoch": 1.5499018032494196,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 6511
    },
    {
      "epoch": 1.5501398559780992,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 6512
    },
    {
      "epoch": 1.5503779087067786,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 6513
    },
    {
      "epoch": 1.550615961435458,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 6514
    },
    {
      "epoch": 1.5508540141641374,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 6515
    },
    {
      "epoch": 1.5510920668928168,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 6516
    },
    {
      "epoch": 1.5513301196214961,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.348,
      "step": 6517
    },
    {
      "epoch": 1.5515681723501755,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 6518
    },
    {
      "epoch": 1.5518062250788551,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 6519
    },
    {
      "epoch": 1.5520442778075343,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 6520
    },
    {
      "epoch": 1.552282330536214,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 6521
    },
    {
      "epoch": 1.552520383264893,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 6522
    },
    {
      "epoch": 1.5527584359935727,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 6523
    },
    {
      "epoch": 1.5529964887222518,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 6524
    },
    {
      "epoch": 1.5532345414509314,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 6525
    },
    {
      "epoch": 1.5534725941796108,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 6526
    },
    {
      "epoch": 1.5537106469082902,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 6527
    },
    {
      "epoch": 1.5539486996369696,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 6528
    },
    {
      "epoch": 1.554186752365649,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.3079,
      "step": 6529
    },
    {
      "epoch": 1.5544248050943283,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 6530
    },
    {
      "epoch": 1.5546628578230077,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 6531
    },
    {
      "epoch": 1.5549009105516873,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3196,
      "step": 6532
    },
    {
      "epoch": 1.5551389632803665,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 6533
    },
    {
      "epoch": 1.555377016009046,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 6534
    },
    {
      "epoch": 1.5556150687377253,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 6535
    },
    {
      "epoch": 1.5558531214664049,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 6536
    },
    {
      "epoch": 1.5560911741950842,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 6537
    },
    {
      "epoch": 1.5563292269237636,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 6538
    },
    {
      "epoch": 1.556567279652443,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 6539
    },
    {
      "epoch": 1.5568053323811224,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 6540
    },
    {
      "epoch": 1.5570433851098018,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4719,
      "step": 6541
    },
    {
      "epoch": 1.5572814378384812,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 6542
    },
    {
      "epoch": 1.5575194905671608,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 6543
    },
    {
      "epoch": 1.55775754329584,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 6544
    },
    {
      "epoch": 1.5579955960245195,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 6545
    },
    {
      "epoch": 1.5582336487531987,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 6546
    },
    {
      "epoch": 1.5584717014818783,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3649,
      "step": 6547
    },
    {
      "epoch": 1.5587097542105577,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 6548
    },
    {
      "epoch": 1.558947806939237,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 6549
    },
    {
      "epoch": 1.5591858596679165,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3872,
      "step": 6550
    },
    {
      "epoch": 1.5594239123965958,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 6551
    },
    {
      "epoch": 1.5596619651252752,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.295,
      "step": 6552
    },
    {
      "epoch": 1.5599000178539546,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 6553
    },
    {
      "epoch": 1.5601380705826342,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 6554
    },
    {
      "epoch": 1.5603761233113134,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 6555
    },
    {
      "epoch": 1.560614176039993,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 6556
    },
    {
      "epoch": 1.5608522287686721,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 6557
    },
    {
      "epoch": 1.5610902814973517,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 6558
    },
    {
      "epoch": 1.561328334226031,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 6559
    },
    {
      "epoch": 1.5615663869547105,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 6560
    },
    {
      "epoch": 1.56180443968339,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 6561
    },
    {
      "epoch": 1.5620424924120693,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7665,
      "step": 6562
    },
    {
      "epoch": 1.5622805451407487,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3031,
      "step": 6563
    },
    {
      "epoch": 1.562518597869428,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 6564
    },
    {
      "epoch": 1.5627566505981076,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9556,
      "step": 6565
    },
    {
      "epoch": 1.5629947033267868,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 6566
    },
    {
      "epoch": 1.5632327560554664,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 6567
    },
    {
      "epoch": 1.5634708087841456,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3773,
      "step": 6568
    },
    {
      "epoch": 1.5637088615128252,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 6569
    },
    {
      "epoch": 1.5639469142415043,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 6570
    },
    {
      "epoch": 1.564184966970184,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 6571
    },
    {
      "epoch": 1.5644230196988633,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 6572
    },
    {
      "epoch": 1.5646610724275427,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 6573
    },
    {
      "epoch": 1.564899125156222,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3455,
      "step": 6574
    },
    {
      "epoch": 1.5651371778849015,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 6575
    },
    {
      "epoch": 1.5653752306135809,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 6576
    },
    {
      "epoch": 1.5656132833422602,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 6577
    },
    {
      "epoch": 1.5658513360709398,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 6578
    },
    {
      "epoch": 1.566089388799619,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 6579
    },
    {
      "epoch": 1.5663274415282986,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 6580
    },
    {
      "epoch": 1.5665654942569778,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 6581
    },
    {
      "epoch": 1.5668035469856574,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 6582
    },
    {
      "epoch": 1.5670415997143368,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3926,
      "step": 6583
    },
    {
      "epoch": 1.5672796524430161,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 6584
    },
    {
      "epoch": 1.5675177051716955,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 6585
    },
    {
      "epoch": 1.567755757900375,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 6586
    },
    {
      "epoch": 1.5679938106290543,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 6587
    },
    {
      "epoch": 1.5682318633577337,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 6588
    },
    {
      "epoch": 1.5684699160864133,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 6589
    },
    {
      "epoch": 1.5687079688150924,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 6590
    },
    {
      "epoch": 1.568946021543772,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 6591
    },
    {
      "epoch": 1.5691840742724512,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 6592
    },
    {
      "epoch": 1.5694221270011308,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 6593
    },
    {
      "epoch": 1.5696601797298102,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 6594
    },
    {
      "epoch": 1.5698982324584896,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 6595
    },
    {
      "epoch": 1.570136285187169,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 6596
    },
    {
      "epoch": 1.5703743379158484,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 6597
    },
    {
      "epoch": 1.5706123906445277,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2923,
      "step": 6598
    },
    {
      "epoch": 1.5708504433732071,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 6599
    },
    {
      "epoch": 1.5710884961018867,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 6600
    },
    {
      "epoch": 1.5710884961018867,
      "eval_runtime": 0.514,
      "eval_samples_per_second": 1.945,
      "eval_steps_per_second": 1.945,
      "step": 6600
    },
    {
      "epoch": 1.5713265488305659,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 6601
    },
    {
      "epoch": 1.5715646015592455,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 6602
    },
    {
      "epoch": 1.5718026542879246,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 6603
    },
    {
      "epoch": 1.5720407070166043,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4731,
      "step": 6604
    },
    {
      "epoch": 1.5722787597452834,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 6605
    },
    {
      "epoch": 1.572516812473963,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 6606
    },
    {
      "epoch": 1.5727548652026424,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 6607
    },
    {
      "epoch": 1.5729929179313218,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 6608
    },
    {
      "epoch": 1.5732309706600012,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 6609
    },
    {
      "epoch": 1.5734690233886806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 6610
    },
    {
      "epoch": 1.5737070761173602,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 6611
    },
    {
      "epoch": 1.5739451288460393,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 6612
    },
    {
      "epoch": 1.574183181574719,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 6613
    },
    {
      "epoch": 1.574421234303398,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 6614
    },
    {
      "epoch": 1.5746592870320777,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 6615
    },
    {
      "epoch": 1.5748973397607569,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 6616
    },
    {
      "epoch": 1.5751353924894365,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 6617
    },
    {
      "epoch": 1.5753734452181158,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.6296,
      "step": 6618
    },
    {
      "epoch": 1.5756114979467952,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 6619
    },
    {
      "epoch": 1.5758495506754746,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 6620
    },
    {
      "epoch": 1.576087603404154,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 6621
    },
    {
      "epoch": 1.5763256561328334,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 6622
    },
    {
      "epoch": 1.5765637088615128,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 6623
    },
    {
      "epoch": 1.5768017615901924,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 6624
    },
    {
      "epoch": 1.5770398143188715,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 6625
    },
    {
      "epoch": 1.5772778670475511,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 6626
    },
    {
      "epoch": 1.5775159197762303,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 6627
    },
    {
      "epoch": 1.57775397250491,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5196,
      "step": 6628
    },
    {
      "epoch": 1.5779920252335893,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 6629
    },
    {
      "epoch": 1.5782300779622687,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 6630
    },
    {
      "epoch": 1.578468130690948,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 6631
    },
    {
      "epoch": 1.5787061834196274,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 6632
    },
    {
      "epoch": 1.5789442361483068,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 6633
    },
    {
      "epoch": 1.5791822888769862,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3558,
      "step": 6634
    },
    {
      "epoch": 1.5794203416056658,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 6635
    },
    {
      "epoch": 1.579658394334345,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 6636
    },
    {
      "epoch": 1.5798964470630246,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 6637
    },
    {
      "epoch": 1.5801344997917037,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2877,
      "step": 6638
    },
    {
      "epoch": 1.5803725525203833,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 6639
    },
    {
      "epoch": 1.5806106052490627,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 6640
    },
    {
      "epoch": 1.580848657977742,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 6641
    },
    {
      "epoch": 1.5810867107064215,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 6642
    },
    {
      "epoch": 1.5813247634351009,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.7325,
      "step": 6643
    },
    {
      "epoch": 1.5815628161637802,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.6501,
      "step": 6644
    },
    {
      "epoch": 1.5818008688924596,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 6645
    },
    {
      "epoch": 1.5820389216211392,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.5384,
      "step": 6646
    },
    {
      "epoch": 1.5822769743498184,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 6647
    },
    {
      "epoch": 1.582515027078498,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.5885,
      "step": 6648
    },
    {
      "epoch": 1.5827530798071772,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 6649
    },
    {
      "epoch": 1.5829911325358568,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 6650
    },
    {
      "epoch": 1.583229185264536,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 6651
    },
    {
      "epoch": 1.5834672379932155,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 6652
    },
    {
      "epoch": 1.583705290721895,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 6653
    },
    {
      "epoch": 1.5839433434505743,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4182,
      "step": 6654
    },
    {
      "epoch": 1.5841813961792537,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 6655
    },
    {
      "epoch": 1.584419448907933,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4468,
      "step": 6656
    },
    {
      "epoch": 1.5846575016366127,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 6657
    },
    {
      "epoch": 1.5848955543652918,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 6658
    },
    {
      "epoch": 1.5851336070939714,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 6659
    },
    {
      "epoch": 1.5853716598226506,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 6660
    },
    {
      "epoch": 1.5856097125513302,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 6661
    },
    {
      "epoch": 1.5858477652800094,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 6662
    },
    {
      "epoch": 1.586085818008689,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 6663
    },
    {
      "epoch": 1.5863238707373684,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.4483,
      "step": 6664
    },
    {
      "epoch": 1.5865619234660477,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3946,
      "step": 6665
    },
    {
      "epoch": 1.5867999761947271,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 6666
    },
    {
      "epoch": 1.5870380289234065,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2526,
      "step": 6667
    },
    {
      "epoch": 1.587276081652086,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4451,
      "step": 6668
    },
    {
      "epoch": 1.5875141343807653,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 6669
    },
    {
      "epoch": 1.5877521871094449,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 6670
    },
    {
      "epoch": 1.587990239838124,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 6671
    },
    {
      "epoch": 1.5882282925668036,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 6672
    },
    {
      "epoch": 1.5884663452954828,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 6673
    },
    {
      "epoch": 1.5887043980241624,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3532,
      "step": 6674
    },
    {
      "epoch": 1.5889424507528418,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 6675
    },
    {
      "epoch": 1.5891805034815212,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2919,
      "step": 6676
    },
    {
      "epoch": 1.5894185562102006,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 6677
    },
    {
      "epoch": 1.58965660893888,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 6678
    },
    {
      "epoch": 1.5898946616675593,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 6679
    },
    {
      "epoch": 1.5901327143962387,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 6680
    },
    {
      "epoch": 1.5903707671249183,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 6681
    },
    {
      "epoch": 1.5906088198535975,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 6682
    },
    {
      "epoch": 1.590846872582277,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 6683
    },
    {
      "epoch": 1.5910849253109562,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 6684
    },
    {
      "epoch": 1.5913229780396358,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 6685
    },
    {
      "epoch": 1.5915610307683152,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.455,
      "step": 6686
    },
    {
      "epoch": 1.5917990834969946,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3425,
      "step": 6687
    },
    {
      "epoch": 1.592037136225674,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 6688
    },
    {
      "epoch": 1.5922751889543534,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 6689
    },
    {
      "epoch": 1.5925132416830328,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4066,
      "step": 6690
    },
    {
      "epoch": 1.5927512944117121,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 6691
    },
    {
      "epoch": 1.5929893471403918,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 6692
    },
    {
      "epoch": 1.593227399869071,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 6693
    },
    {
      "epoch": 1.5934654525977505,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 6694
    },
    {
      "epoch": 1.5937035053264297,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 6695
    },
    {
      "epoch": 1.5939415580551093,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 6696
    },
    {
      "epoch": 1.5941796107837884,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3898,
      "step": 6697
    },
    {
      "epoch": 1.594417663512468,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6848,
      "step": 6698
    },
    {
      "epoch": 1.5946557162411474,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 6699
    },
    {
      "epoch": 1.5948937689698268,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 6700
    },
    {
      "epoch": 1.5948937689698268,
      "eval_runtime": 0.5221,
      "eval_samples_per_second": 1.915,
      "eval_steps_per_second": 1.915,
      "step": 6700
    },
    {
      "epoch": 1.5951318216985062,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3767,
      "step": 6701
    },
    {
      "epoch": 1.5953698744271856,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 6702
    },
    {
      "epoch": 1.595607927155865,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 6703
    },
    {
      "epoch": 1.5958459798845444,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 6704
    },
    {
      "epoch": 1.596084032613224,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 6705
    },
    {
      "epoch": 1.5963220853419031,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 6706
    },
    {
      "epoch": 1.5965601380705827,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 6707
    },
    {
      "epoch": 1.5967981907992619,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 6708
    },
    {
      "epoch": 1.5970362435279415,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 6709
    },
    {
      "epoch": 1.5972742962566209,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.304,
      "step": 6710
    },
    {
      "epoch": 1.5975123489853003,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 6711
    },
    {
      "epoch": 1.5977504017139796,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 6712
    },
    {
      "epoch": 1.597988454442659,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 6713
    },
    {
      "epoch": 1.5982265071713384,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3934,
      "step": 6714
    },
    {
      "epoch": 1.5984645599000178,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 6715
    },
    {
      "epoch": 1.5987026126286974,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 6716
    },
    {
      "epoch": 1.5989406653573766,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 6717
    },
    {
      "epoch": 1.5991787180860562,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 6718
    },
    {
      "epoch": 1.5994167708147353,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 6719
    },
    {
      "epoch": 1.599654823543415,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 6720
    },
    {
      "epoch": 1.5998928762720943,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5202,
      "step": 6721
    },
    {
      "epoch": 1.6001309290007737,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 6722
    },
    {
      "epoch": 1.600368981729453,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 6723
    },
    {
      "epoch": 1.6006070344581325,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 6724
    },
    {
      "epoch": 1.6008450871868118,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 6725
    },
    {
      "epoch": 1.6010831399154912,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3275,
      "step": 6726
    },
    {
      "epoch": 1.6013211926441708,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 6727
    },
    {
      "epoch": 1.60155924537285,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 6728
    },
    {
      "epoch": 1.6017972981015296,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 6729
    },
    {
      "epoch": 1.6020353508302088,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 6730
    },
    {
      "epoch": 1.6022734035588884,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3235,
      "step": 6731
    },
    {
      "epoch": 1.6025114562875677,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 6732
    },
    {
      "epoch": 1.6027495090162471,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4589,
      "step": 6733
    },
    {
      "epoch": 1.6029875617449265,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 6734
    },
    {
      "epoch": 1.603225614473606,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3327,
      "step": 6735
    },
    {
      "epoch": 1.6034636672022853,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 6736
    },
    {
      "epoch": 1.6037017199309647,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 6737
    },
    {
      "epoch": 1.6039397726596443,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 6738
    },
    {
      "epoch": 1.6041778253883234,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 6739
    },
    {
      "epoch": 1.604415878117003,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 6740
    },
    {
      "epoch": 1.6046539308456822,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.6839,
      "step": 6741
    },
    {
      "epoch": 1.6048919835743618,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3342,
      "step": 6742
    },
    {
      "epoch": 1.605130036303041,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 6743
    },
    {
      "epoch": 1.6053680890317206,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 6744
    },
    {
      "epoch": 1.6056061417604,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3515,
      "step": 6745
    },
    {
      "epoch": 1.6058441944890793,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 6746
    },
    {
      "epoch": 1.6060822472177587,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 6747
    },
    {
      "epoch": 1.606320299946438,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 6748
    },
    {
      "epoch": 1.6065583526751175,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 6749
    },
    {
      "epoch": 1.6067964054037969,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4606,
      "step": 6750
    },
    {
      "epoch": 1.6070344581324765,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 6751
    },
    {
      "epoch": 1.6072725108611556,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 6752
    },
    {
      "epoch": 1.6075105635898352,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 6753
    },
    {
      "epoch": 1.6077486163185144,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 6754
    },
    {
      "epoch": 1.607986669047194,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3953,
      "step": 6755
    },
    {
      "epoch": 1.6082247217758734,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 6756
    },
    {
      "epoch": 1.6084627745045528,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 6757
    },
    {
      "epoch": 1.6087008272332322,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4209,
      "step": 6758
    },
    {
      "epoch": 1.6089388799619115,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 6759
    },
    {
      "epoch": 1.609176932690591,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 6760
    },
    {
      "epoch": 1.6094149854192703,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 6761
    },
    {
      "epoch": 1.60965303814795,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 6762
    },
    {
      "epoch": 1.609891090876629,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 6763
    },
    {
      "epoch": 1.6101291436053087,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 6764
    },
    {
      "epoch": 1.6103671963339878,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 6765
    },
    {
      "epoch": 1.6106052490626674,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 6766
    },
    {
      "epoch": 1.6108433017913468,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 6767
    },
    {
      "epoch": 1.6110813545200262,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 6768
    },
    {
      "epoch": 1.6113194072487056,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3772,
      "step": 6769
    },
    {
      "epoch": 1.611557459977385,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 6770
    },
    {
      "epoch": 1.6117955127060644,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4325,
      "step": 6771
    },
    {
      "epoch": 1.6120335654347437,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 6772
    },
    {
      "epoch": 1.6122716181634233,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 6773
    },
    {
      "epoch": 1.6125096708921025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 6774
    },
    {
      "epoch": 1.6127477236207821,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 6775
    },
    {
      "epoch": 1.6129857763494613,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 6776
    },
    {
      "epoch": 1.6132238290781409,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.268,
      "step": 6777
    },
    {
      "epoch": 1.6134618818068203,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4403,
      "step": 6778
    },
    {
      "epoch": 1.6136999345354996,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 6779
    },
    {
      "epoch": 1.613937987264179,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 6780
    },
    {
      "epoch": 1.6141760399928584,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 6781
    },
    {
      "epoch": 1.6144140927215378,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2518,
      "step": 6782
    },
    {
      "epoch": 1.6146521454502172,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 6783
    },
    {
      "epoch": 1.6148901981788968,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 6784
    },
    {
      "epoch": 1.615128250907576,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4826,
      "step": 6785
    },
    {
      "epoch": 1.6153663036362556,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 6786
    },
    {
      "epoch": 1.6156043563649347,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 6787
    },
    {
      "epoch": 1.6158424090936143,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 6788
    },
    {
      "epoch": 1.6160804618222935,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4176,
      "step": 6789
    },
    {
      "epoch": 1.616318514550973,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 6790
    },
    {
      "epoch": 1.6165565672796525,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3356,
      "step": 6791
    },
    {
      "epoch": 1.6167946200083319,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 6792
    },
    {
      "epoch": 1.6170326727370112,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 6793
    },
    {
      "epoch": 1.6172707254656906,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 6794
    },
    {
      "epoch": 1.61750877819437,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 6795
    },
    {
      "epoch": 1.6177468309230494,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 6796
    },
    {
      "epoch": 1.617984883651729,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5178,
      "step": 6797
    },
    {
      "epoch": 1.6182229363804082,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 6798
    },
    {
      "epoch": 1.6184609891090878,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.436,
      "step": 6799
    },
    {
      "epoch": 1.618699041837767,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.235,
      "step": 6800
    },
    {
      "epoch": 1.618699041837767,
      "eval_runtime": 0.5148,
      "eval_samples_per_second": 1.942,
      "eval_steps_per_second": 1.942,
      "step": 6800
    },
    {
      "epoch": 1.6189370945664465,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 6801
    },
    {
      "epoch": 1.619175147295126,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 6802
    },
    {
      "epoch": 1.6194132000238053,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 6803
    },
    {
      "epoch": 1.6196512527524847,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 6804
    },
    {
      "epoch": 1.619889305481164,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 6805
    },
    {
      "epoch": 1.6201273582098434,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 6806
    },
    {
      "epoch": 1.6203654109385228,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 6807
    },
    {
      "epoch": 1.6206034636672024,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 6808
    },
    {
      "epoch": 1.6208415163958816,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 6809
    },
    {
      "epoch": 1.6210795691245612,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 6810
    },
    {
      "epoch": 1.6213176218532404,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2432,
      "step": 6811
    },
    {
      "epoch": 1.62155567458192,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.344,
      "step": 6812
    },
    {
      "epoch": 1.6217937273105993,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3395,
      "step": 6813
    },
    {
      "epoch": 1.6220317800392787,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 6814
    },
    {
      "epoch": 1.622269832767958,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 6815
    },
    {
      "epoch": 1.6225078854966375,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 6816
    },
    {
      "epoch": 1.6227459382253169,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 6817
    },
    {
      "epoch": 1.6229839909539963,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3833,
      "step": 6818
    },
    {
      "epoch": 1.6232220436826759,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 6819
    },
    {
      "epoch": 1.623460096411355,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 6820
    },
    {
      "epoch": 1.6236981491400346,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 6821
    },
    {
      "epoch": 1.6239362018687138,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 6822
    },
    {
      "epoch": 1.6241742545973934,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 6823
    },
    {
      "epoch": 1.6244123073260728,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 6824
    },
    {
      "epoch": 1.6246503600547522,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3176,
      "step": 6825
    },
    {
      "epoch": 1.6248884127834315,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 6826
    },
    {
      "epoch": 1.625126465512111,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 6827
    },
    {
      "epoch": 1.6253645182407903,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3525,
      "step": 6828
    },
    {
      "epoch": 1.6256025709694697,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4312,
      "step": 6829
    },
    {
      "epoch": 1.6258406236981493,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 6830
    },
    {
      "epoch": 1.6260786764268285,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 6831
    },
    {
      "epoch": 1.626316729155508,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3656,
      "step": 6832
    },
    {
      "epoch": 1.6265547818841872,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3982,
      "step": 6833
    },
    {
      "epoch": 1.6267928346128668,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 6834
    },
    {
      "epoch": 1.627030887341546,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 6835
    },
    {
      "epoch": 1.6272689400702256,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4125,
      "step": 6836
    },
    {
      "epoch": 1.627506992798905,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 6837
    },
    {
      "epoch": 1.6277450455275844,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 6838
    },
    {
      "epoch": 1.6279830982562637,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 6839
    },
    {
      "epoch": 1.6282211509849431,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 6840
    },
    {
      "epoch": 1.6284592037136225,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 6841
    },
    {
      "epoch": 1.628697256442302,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 6842
    },
    {
      "epoch": 1.6289353091709815,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 6843
    },
    {
      "epoch": 1.6291733618996607,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 6844
    },
    {
      "epoch": 1.6294114146283403,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 6845
    },
    {
      "epoch": 1.6296494673570194,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 6846
    },
    {
      "epoch": 1.629887520085699,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 6847
    },
    {
      "epoch": 1.6301255728143784,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 6848
    },
    {
      "epoch": 1.6303636255430578,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3025,
      "step": 6849
    },
    {
      "epoch": 1.6306016782717372,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 6850
    },
    {
      "epoch": 1.6308397310004166,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 6851
    },
    {
      "epoch": 1.631077783729096,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 6852
    },
    {
      "epoch": 1.6313158364577753,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 6853
    },
    {
      "epoch": 1.631553889186455,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 6854
    },
    {
      "epoch": 1.631791941915134,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 6855
    },
    {
      "epoch": 1.6320299946438137,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3442,
      "step": 6856
    },
    {
      "epoch": 1.6322680473724929,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 6857
    },
    {
      "epoch": 1.6325061001011725,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.385,
      "step": 6858
    },
    {
      "epoch": 1.6327441528298519,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 6859
    },
    {
      "epoch": 1.6329822055585312,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 6860
    },
    {
      "epoch": 1.6332202582872106,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 6861
    },
    {
      "epoch": 1.63345831101589,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2572,
      "step": 6862
    },
    {
      "epoch": 1.6336963637445694,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 6863
    },
    {
      "epoch": 1.6339344164732488,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4593,
      "step": 6864
    },
    {
      "epoch": 1.6341724692019284,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 6865
    },
    {
      "epoch": 1.6344105219306075,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 6866
    },
    {
      "epoch": 1.6346485746592871,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 6867
    },
    {
      "epoch": 1.6348866273879663,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 6868
    },
    {
      "epoch": 1.635124680116646,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 6869
    },
    {
      "epoch": 1.635362732845325,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 6870
    },
    {
      "epoch": 1.6356007855740047,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 6871
    },
    {
      "epoch": 1.635838838302684,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 6872
    },
    {
      "epoch": 1.6360768910313634,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 6873
    },
    {
      "epoch": 1.6363149437600428,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 6874
    },
    {
      "epoch": 1.6365529964887222,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 6875
    },
    {
      "epoch": 1.6367910492174018,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 6876
    },
    {
      "epoch": 1.637029101946081,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4667,
      "step": 6877
    },
    {
      "epoch": 1.6372671546747606,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 6878
    },
    {
      "epoch": 1.6375052074034397,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 6879
    },
    {
      "epoch": 1.6377432601321193,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 6880
    },
    {
      "epoch": 1.6379813128607985,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 6881
    },
    {
      "epoch": 1.6382193655894781,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 6882
    },
    {
      "epoch": 1.6384574183181575,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.6399,
      "step": 6883
    },
    {
      "epoch": 1.6386954710468369,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 6884
    },
    {
      "epoch": 1.6389335237755163,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 6885
    },
    {
      "epoch": 1.6391715765041956,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5753,
      "step": 6886
    },
    {
      "epoch": 1.639409629232875,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 6887
    },
    {
      "epoch": 1.6396476819615544,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 6888
    },
    {
      "epoch": 1.639885734690234,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4141,
      "step": 6889
    },
    {
      "epoch": 1.6401237874189132,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 6890
    },
    {
      "epoch": 1.6403618401475928,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 6891
    },
    {
      "epoch": 1.640599892876272,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 6892
    },
    {
      "epoch": 1.6408379456049516,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 6893
    },
    {
      "epoch": 1.641075998333631,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.5873,
      "step": 6894
    },
    {
      "epoch": 1.6413140510623103,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 6895
    },
    {
      "epoch": 1.6415521037909897,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 6896
    },
    {
      "epoch": 1.641790156519669,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 6897
    },
    {
      "epoch": 1.6420282092483485,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 6898
    },
    {
      "epoch": 1.6422662619770279,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 6899
    },
    {
      "epoch": 1.6425043147057075,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 6900
    },
    {
      "epoch": 1.6425043147057075,
      "eval_runtime": 0.5079,
      "eval_samples_per_second": 1.969,
      "eval_steps_per_second": 1.969,
      "step": 6900
    },
    {
      "epoch": 1.6427423674343866,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 6901
    },
    {
      "epoch": 1.6429804201630662,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 6902
    },
    {
      "epoch": 1.6432184728917454,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 6903
    },
    {
      "epoch": 1.643456525620425,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 6904
    },
    {
      "epoch": 1.6436945783491044,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 6905
    },
    {
      "epoch": 1.6439326310777838,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 6906
    },
    {
      "epoch": 1.6441706838064631,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 6907
    },
    {
      "epoch": 1.6444087365351425,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 6908
    },
    {
      "epoch": 1.644646789263822,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 6909
    },
    {
      "epoch": 1.6448848419925013,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 6910
    },
    {
      "epoch": 1.645122894721181,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 6911
    },
    {
      "epoch": 1.64536094744986,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 6912
    },
    {
      "epoch": 1.6455990001785397,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 6913
    },
    {
      "epoch": 1.6458370529072188,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5198,
      "step": 6914
    },
    {
      "epoch": 1.6460751056358984,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4001,
      "step": 6915
    },
    {
      "epoch": 1.6463131583645776,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 6916
    },
    {
      "epoch": 1.6465512110932572,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 6917
    },
    {
      "epoch": 1.6467892638219366,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 6918
    },
    {
      "epoch": 1.647027316550616,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 6919
    },
    {
      "epoch": 1.6472653692792953,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3021,
      "step": 6920
    },
    {
      "epoch": 1.6475034220079747,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 6921
    },
    {
      "epoch": 1.6477414747366543,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 6922
    },
    {
      "epoch": 1.6479795274653335,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 6923
    },
    {
      "epoch": 1.648217580194013,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 6924
    },
    {
      "epoch": 1.6484556329226923,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 6925
    },
    {
      "epoch": 1.6486936856513719,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 6926
    },
    {
      "epoch": 1.648931738380051,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 6927
    },
    {
      "epoch": 1.6491697911087306,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3002,
      "step": 6928
    },
    {
      "epoch": 1.64940784383741,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 6929
    },
    {
      "epoch": 1.6496458965660894,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 6930
    },
    {
      "epoch": 1.6498839492947688,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 6931
    },
    {
      "epoch": 1.6501220020234482,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5102,
      "step": 6932
    },
    {
      "epoch": 1.6503600547521275,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 6933
    },
    {
      "epoch": 1.650598107480807,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 6934
    },
    {
      "epoch": 1.6508361602094865,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 6935
    },
    {
      "epoch": 1.6510742129381657,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 6936
    },
    {
      "epoch": 1.6513122656668453,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 6937
    },
    {
      "epoch": 1.6515503183955245,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 6938
    },
    {
      "epoch": 1.651788371124204,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 6939
    },
    {
      "epoch": 1.6520264238528835,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5052,
      "step": 6940
    },
    {
      "epoch": 1.6522644765815628,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 6941
    },
    {
      "epoch": 1.6525025293102422,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 6942
    },
    {
      "epoch": 1.6527405820389216,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 6943
    },
    {
      "epoch": 1.652978634767601,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 6944
    },
    {
      "epoch": 1.6532166874962804,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 6945
    },
    {
      "epoch": 1.65345474022496,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 6946
    },
    {
      "epoch": 1.6536927929536391,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 6947
    },
    {
      "epoch": 1.6539308456823187,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 6948
    },
    {
      "epoch": 1.654168898410998,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 6949
    },
    {
      "epoch": 1.6544069511396775,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 6950
    },
    {
      "epoch": 1.654645003868357,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 6951
    },
    {
      "epoch": 1.6548830565970363,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 6952
    },
    {
      "epoch": 1.6551211093257157,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 6953
    },
    {
      "epoch": 1.655359162054395,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 6954
    },
    {
      "epoch": 1.6555972147830744,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 6955
    },
    {
      "epoch": 1.6558352675117538,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 6956
    },
    {
      "epoch": 1.6560733202404334,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 6957
    },
    {
      "epoch": 1.6563113729691126,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 6958
    },
    {
      "epoch": 1.6565494256977922,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 6959
    },
    {
      "epoch": 1.6567874784264713,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 6960
    },
    {
      "epoch": 1.657025531155151,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 6961
    },
    {
      "epoch": 1.65726358388383,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 6962
    },
    {
      "epoch": 1.6575016366125097,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 6963
    },
    {
      "epoch": 1.657739689341189,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4251,
      "step": 6964
    },
    {
      "epoch": 1.6579777420698685,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 6965
    },
    {
      "epoch": 1.6582157947985479,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.5134,
      "step": 6966
    },
    {
      "epoch": 1.6584538475272272,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 6967
    },
    {
      "epoch": 1.6586919002559068,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 6968
    },
    {
      "epoch": 1.658929952984586,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 6969
    },
    {
      "epoch": 1.6591680057132656,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 6970
    },
    {
      "epoch": 1.6594060584419448,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 6971
    },
    {
      "epoch": 1.6596441111706244,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 6972
    },
    {
      "epoch": 1.6598821638993035,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 6973
    },
    {
      "epoch": 1.6601202166279831,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 6974
    },
    {
      "epoch": 1.6603582693566625,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 6975
    },
    {
      "epoch": 1.660596322085342,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 6976
    },
    {
      "epoch": 1.6608343748140213,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 6977
    },
    {
      "epoch": 1.6610724275427007,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2798,
      "step": 6978
    },
    {
      "epoch": 1.66131048027138,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 6979
    },
    {
      "epoch": 1.6615485330000594,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 6980
    },
    {
      "epoch": 1.661786585728739,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 6981
    },
    {
      "epoch": 1.6620246384574182,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 6982
    },
    {
      "epoch": 1.6622626911860978,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.585,
      "step": 6983
    },
    {
      "epoch": 1.662500743914777,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 6984
    },
    {
      "epoch": 1.6627387966434566,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 6985
    },
    {
      "epoch": 1.662976849372136,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 6986
    },
    {
      "epoch": 1.6632149021008154,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 6987
    },
    {
      "epoch": 1.6634529548294947,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 6988
    },
    {
      "epoch": 1.6636910075581741,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 6989
    },
    {
      "epoch": 1.6639290602868535,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 6990
    },
    {
      "epoch": 1.6641671130155329,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3707,
      "step": 6991
    },
    {
      "epoch": 1.6644051657442125,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 6992
    },
    {
      "epoch": 1.6646432184728917,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 6993
    },
    {
      "epoch": 1.6648812712015713,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 6994
    },
    {
      "epoch": 1.6651193239302504,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 6995
    },
    {
      "epoch": 1.66535737665893,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3064,
      "step": 6996
    },
    {
      "epoch": 1.6655954293876094,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4958,
      "step": 6997
    },
    {
      "epoch": 1.6658334821162888,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 6998
    },
    {
      "epoch": 1.6660715348449682,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 6999
    },
    {
      "epoch": 1.6663095875736476,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4916,
      "step": 7000
    },
    {
      "epoch": 1.6663095875736476,
      "eval_runtime": 0.5064,
      "eval_samples_per_second": 1.975,
      "eval_steps_per_second": 1.975,
      "step": 7000
    },
    {
      "epoch": 1.6667856930310063,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 7001
    },
    {
      "epoch": 1.667023745759686,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 7002
    },
    {
      "epoch": 1.667261798488365,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 7003
    },
    {
      "epoch": 1.6674998512170447,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 7004
    },
    {
      "epoch": 1.6677379039457239,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 7005
    },
    {
      "epoch": 1.6679759566744035,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 7006
    },
    {
      "epoch": 1.6682140094030826,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 7007
    },
    {
      "epoch": 1.6684520621317622,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 7008
    },
    {
      "epoch": 1.6686901148604416,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 7009
    },
    {
      "epoch": 1.668928167589121,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 7010
    },
    {
      "epoch": 1.6691662203178004,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 7011
    },
    {
      "epoch": 1.6694042730464798,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 7012
    },
    {
      "epoch": 1.6696423257751594,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 7013
    },
    {
      "epoch": 1.6698803785038385,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 7014
    },
    {
      "epoch": 1.6701184312325181,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 7015
    },
    {
      "epoch": 1.6703564839611973,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 7016
    },
    {
      "epoch": 1.670594536689877,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 7017
    },
    {
      "epoch": 1.670832589418556,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4187,
      "step": 7018
    },
    {
      "epoch": 1.6710706421472357,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 7019
    },
    {
      "epoch": 1.671308694875915,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 7020
    },
    {
      "epoch": 1.6715467476045944,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 7021
    },
    {
      "epoch": 1.6717848003332738,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 7022
    },
    {
      "epoch": 1.6720228530619532,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 7023
    },
    {
      "epoch": 1.6722609057906326,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 7024
    },
    {
      "epoch": 1.672498958519312,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3212,
      "step": 7025
    },
    {
      "epoch": 1.6727370112479916,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.5819,
      "step": 7026
    },
    {
      "epoch": 1.6729750639766707,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 7027
    },
    {
      "epoch": 1.6732131167053503,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 7028
    },
    {
      "epoch": 1.6734511694340295,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 7029
    },
    {
      "epoch": 1.673689222162709,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 7030
    },
    {
      "epoch": 1.6739272748913885,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4363,
      "step": 7031
    },
    {
      "epoch": 1.6741653276200679,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 7032
    },
    {
      "epoch": 1.6744033803487473,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 7033
    },
    {
      "epoch": 1.6746414330774266,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 7034
    },
    {
      "epoch": 1.674879485806106,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3385,
      "step": 7035
    },
    {
      "epoch": 1.6751175385347854,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 7036
    },
    {
      "epoch": 1.675355591263465,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 7037
    },
    {
      "epoch": 1.6755936439921442,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 7038
    },
    {
      "epoch": 1.6758316967208238,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 7039
    },
    {
      "epoch": 1.676069749449503,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 7040
    },
    {
      "epoch": 1.6763078021781825,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 7041
    },
    {
      "epoch": 1.676545854906862,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 7042
    },
    {
      "epoch": 1.6767839076355413,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 7043
    },
    {
      "epoch": 1.6770219603642207,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 7044
    },
    {
      "epoch": 1.6772600130929,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 7045
    },
    {
      "epoch": 1.6774980658215795,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 7046
    },
    {
      "epoch": 1.6777361185502588,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 7047
    },
    {
      "epoch": 1.6779741712789384,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.4956,
      "step": 7048
    },
    {
      "epoch": 1.6782122240076176,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 7049
    },
    {
      "epoch": 1.6784502767362972,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 7050
    },
    {
      "epoch": 1.6786883294649764,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 7051
    },
    {
      "epoch": 1.678926382193656,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 7052
    },
    {
      "epoch": 1.6791644349223351,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 7053
    },
    {
      "epoch": 1.6794024876510147,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 7054
    },
    {
      "epoch": 1.6796405403796941,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 7055
    },
    {
      "epoch": 1.6798785931083735,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 7056
    },
    {
      "epoch": 1.680116645837053,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 7057
    },
    {
      "epoch": 1.6803546985657323,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 7058
    },
    {
      "epoch": 1.6805927512944117,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 7059
    },
    {
      "epoch": 1.680830804023091,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 7060
    },
    {
      "epoch": 1.6810688567517706,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.5316,
      "step": 7061
    },
    {
      "epoch": 1.6813069094804498,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 7062
    },
    {
      "epoch": 1.6815449622091294,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 7063
    },
    {
      "epoch": 1.6817830149378086,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 7064
    },
    {
      "epoch": 1.6820210676664882,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 7065
    },
    {
      "epoch": 1.6822591203951676,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 7066
    },
    {
      "epoch": 1.682497173123847,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 7067
    },
    {
      "epoch": 1.6827352258525263,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 7068
    },
    {
      "epoch": 1.6829732785812057,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 7069
    },
    {
      "epoch": 1.683211331309885,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4274,
      "step": 7070
    },
    {
      "epoch": 1.6834493840385645,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3871,
      "step": 7071
    },
    {
      "epoch": 1.683687436767244,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 7072
    },
    {
      "epoch": 1.6839254894959232,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 7073
    },
    {
      "epoch": 1.6841635422246028,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.43,
      "step": 7074
    },
    {
      "epoch": 1.684401594953282,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3951,
      "step": 7075
    },
    {
      "epoch": 1.6846396476819616,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 7076
    },
    {
      "epoch": 1.684877700410641,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 7077
    },
    {
      "epoch": 1.6851157531393204,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 7078
    },
    {
      "epoch": 1.6853538058679998,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 7079
    },
    {
      "epoch": 1.6855918585966791,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 7080
    },
    {
      "epoch": 1.6858299113253585,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.5218,
      "step": 7081
    },
    {
      "epoch": 1.686067964054038,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 7082
    },
    {
      "epoch": 1.6863060167827175,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 7083
    },
    {
      "epoch": 1.6865440695113967,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 7084
    },
    {
      "epoch": 1.6867821222400763,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 7085
    },
    {
      "epoch": 1.6870201749687554,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4371,
      "step": 7086
    },
    {
      "epoch": 1.687258227697435,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 7087
    },
    {
      "epoch": 1.6874962804261144,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 7088
    },
    {
      "epoch": 1.6877343331547938,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 7089
    },
    {
      "epoch": 1.6879723858834732,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8466,
      "step": 7090
    },
    {
      "epoch": 1.6882104386121526,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 7091
    },
    {
      "epoch": 1.688448491340832,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 7092
    },
    {
      "epoch": 1.6886865440695114,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 7093
    },
    {
      "epoch": 1.688924596798191,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3463,
      "step": 7094
    },
    {
      "epoch": 1.6891626495268701,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 7095
    },
    {
      "epoch": 1.6894007022555497,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 7096
    },
    {
      "epoch": 1.6896387549842289,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 7097
    },
    {
      "epoch": 1.6898768077129085,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 7098
    },
    {
      "epoch": 1.6901148604415877,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 7099
    },
    {
      "epoch": 1.6903529131702673,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 7100
    },
    {
      "epoch": 1.6903529131702673,
      "eval_runtime": 0.5392,
      "eval_samples_per_second": 1.855,
      "eval_steps_per_second": 1.855,
      "step": 7100
    },
    {
      "epoch": 1.6905909658989466,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 7101
    },
    {
      "epoch": 1.690829018627626,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 7102
    },
    {
      "epoch": 1.6910670713563054,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3891,
      "step": 7103
    },
    {
      "epoch": 1.6913051240849848,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 7104
    },
    {
      "epoch": 1.6915431768136642,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.4345,
      "step": 7105
    },
    {
      "epoch": 1.6917812295423436,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3488,
      "step": 7106
    },
    {
      "epoch": 1.6920192822710232,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4143,
      "step": 7107
    },
    {
      "epoch": 1.6922573349997023,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 7108
    },
    {
      "epoch": 1.692495387728382,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 7109
    },
    {
      "epoch": 1.692733440457061,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 7110
    },
    {
      "epoch": 1.6929714931857407,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 7111
    },
    {
      "epoch": 1.69320954591442,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2895,
      "step": 7112
    },
    {
      "epoch": 1.6934475986430995,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 7113
    },
    {
      "epoch": 1.6936856513717788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 7114
    },
    {
      "epoch": 1.6939237041004582,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3578,
      "step": 7115
    },
    {
      "epoch": 1.6941617568291376,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 7116
    },
    {
      "epoch": 1.694399809557817,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 7117
    },
    {
      "epoch": 1.6946378622864966,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 7118
    },
    {
      "epoch": 1.6948759150151758,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.6049,
      "step": 7119
    },
    {
      "epoch": 1.6951139677438554,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 7120
    },
    {
      "epoch": 1.6953520204725345,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 7121
    },
    {
      "epoch": 1.6955900732012141,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 7122
    },
    {
      "epoch": 1.6958281259298935,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 7123
    },
    {
      "epoch": 1.696066178658573,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 7124
    },
    {
      "epoch": 1.6963042313872523,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 7125
    },
    {
      "epoch": 1.6965422841159317,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 7126
    },
    {
      "epoch": 1.696780336844611,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 7127
    },
    {
      "epoch": 1.6970183895732904,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 7128
    },
    {
      "epoch": 1.69725644230197,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 7129
    },
    {
      "epoch": 1.6974944950306492,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 7130
    },
    {
      "epoch": 1.6977325477593288,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 7131
    },
    {
      "epoch": 1.697970600488008,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 7132
    },
    {
      "epoch": 1.6982086532166876,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 7133
    },
    {
      "epoch": 1.698446705945367,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.5119,
      "step": 7134
    },
    {
      "epoch": 1.6986847586740463,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 7135
    },
    {
      "epoch": 1.6989228114027257,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3625,
      "step": 7136
    },
    {
      "epoch": 1.699160864131405,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 7137
    },
    {
      "epoch": 1.6993989168600845,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 7138
    },
    {
      "epoch": 1.6996369695887639,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 7139
    },
    {
      "epoch": 1.6998750223174435,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 7140
    },
    {
      "epoch": 1.7001130750461226,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 7141
    },
    {
      "epoch": 1.7003511277748022,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 7142
    },
    {
      "epoch": 1.7005891805034814,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3683,
      "step": 7143
    },
    {
      "epoch": 1.700827233232161,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.5339,
      "step": 7144
    },
    {
      "epoch": 1.7010652859608402,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 7145
    },
    {
      "epoch": 1.7013033386895198,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 7146
    },
    {
      "epoch": 1.7015413914181992,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 7147
    },
    {
      "epoch": 1.7017794441468785,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 7148
    },
    {
      "epoch": 1.702017496875558,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 7149
    },
    {
      "epoch": 1.7022555496042373,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 7150
    },
    {
      "epoch": 1.7024936023329167,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4122,
      "step": 7151
    },
    {
      "epoch": 1.702731655061596,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 7152
    },
    {
      "epoch": 1.7029697077902757,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 7153
    },
    {
      "epoch": 1.7032077605189548,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 7154
    },
    {
      "epoch": 1.7034458132476344,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 7155
    },
    {
      "epoch": 1.7036838659763136,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 7156
    },
    {
      "epoch": 1.7039219187049932,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 7157
    },
    {
      "epoch": 1.7041599714336726,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3398,
      "step": 7158
    },
    {
      "epoch": 1.704398024162352,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 7159
    },
    {
      "epoch": 1.7046360768910314,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 7160
    },
    {
      "epoch": 1.7048741296197107,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 7161
    },
    {
      "epoch": 1.7051121823483901,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1956,
      "step": 7162
    },
    {
      "epoch": 1.7053502350770695,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 7163
    },
    {
      "epoch": 1.7055882878057491,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 7164
    },
    {
      "epoch": 1.7058263405344283,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 7165
    },
    {
      "epoch": 1.7060643932631079,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.5481,
      "step": 7166
    },
    {
      "epoch": 1.706302445991787,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5255,
      "step": 7167
    },
    {
      "epoch": 1.7065404987204666,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2931,
      "step": 7168
    },
    {
      "epoch": 1.706778551449146,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 7169
    },
    {
      "epoch": 1.7070166041778254,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 7170
    },
    {
      "epoch": 1.7072546569065048,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 7171
    },
    {
      "epoch": 1.7074927096351842,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 7172
    },
    {
      "epoch": 1.7077307623638636,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 7173
    },
    {
      "epoch": 1.707968815092543,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 7174
    },
    {
      "epoch": 1.7082068678212226,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 7175
    },
    {
      "epoch": 1.7084449205499017,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3799,
      "step": 7176
    },
    {
      "epoch": 1.7086829732785813,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.6198,
      "step": 7177
    },
    {
      "epoch": 1.7089210260072605,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 7178
    },
    {
      "epoch": 1.70915907873594,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 7179
    },
    {
      "epoch": 1.7093971314646195,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4151,
      "step": 7180
    },
    {
      "epoch": 1.7096351841932989,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.6399,
      "step": 7181
    },
    {
      "epoch": 1.7098732369219782,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 7182
    },
    {
      "epoch": 1.7101112896506576,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 7183
    },
    {
      "epoch": 1.710349342379337,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 7184
    },
    {
      "epoch": 1.7105873951080164,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3949,
      "step": 7185
    },
    {
      "epoch": 1.710825447836696,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 7186
    },
    {
      "epoch": 1.7110635005653752,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 7187
    },
    {
      "epoch": 1.7113015532940548,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 7188
    },
    {
      "epoch": 1.711539606022734,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 7189
    },
    {
      "epoch": 1.7117776587514135,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.4356,
      "step": 7190
    },
    {
      "epoch": 1.7120157114800927,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 7191
    },
    {
      "epoch": 1.7122537642087723,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 7192
    },
    {
      "epoch": 1.7124918169374517,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 7193
    },
    {
      "epoch": 1.712729869666131,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 7194
    },
    {
      "epoch": 1.7129679223948104,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 7195
    },
    {
      "epoch": 1.7132059751234898,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 7196
    },
    {
      "epoch": 1.7134440278521692,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 7197
    },
    {
      "epoch": 1.7136820805808486,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 7198
    },
    {
      "epoch": 1.7139201333095282,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 7199
    },
    {
      "epoch": 1.7141581860382074,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 7200
    },
    {
      "epoch": 1.7141581860382074,
      "eval_runtime": 0.5446,
      "eval_samples_per_second": 1.836,
      "eval_steps_per_second": 1.836,
      "step": 7200
    },
    {
      "epoch": 1.714396238766887,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 7201
    },
    {
      "epoch": 1.7146342914955661,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 7202
    },
    {
      "epoch": 1.7148723442242457,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3169,
      "step": 7203
    },
    {
      "epoch": 1.715110396952925,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 7204
    },
    {
      "epoch": 1.7153484496816045,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3653,
      "step": 7205
    },
    {
      "epoch": 1.7155865024102839,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 7206
    },
    {
      "epoch": 1.7158245551389633,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4002,
      "step": 7207
    },
    {
      "epoch": 1.7160626078676426,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 7208
    },
    {
      "epoch": 1.716300660596322,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 7209
    },
    {
      "epoch": 1.7165387133250016,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 7210
    },
    {
      "epoch": 1.7167767660536808,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 7211
    },
    {
      "epoch": 1.7170148187823604,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 7212
    },
    {
      "epoch": 1.7172528715110396,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 7213
    },
    {
      "epoch": 1.7174909242397192,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 7214
    },
    {
      "epoch": 1.7177289769683985,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2327,
      "step": 7215
    },
    {
      "epoch": 1.717967029697078,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 7216
    },
    {
      "epoch": 1.7182050824257573,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 7217
    },
    {
      "epoch": 1.7184431351544367,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 7218
    },
    {
      "epoch": 1.718681187883116,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2957,
      "step": 7219
    },
    {
      "epoch": 1.7189192406117955,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4899,
      "step": 7220
    },
    {
      "epoch": 1.719157293340475,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 7221
    },
    {
      "epoch": 1.7193953460691542,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 7222
    },
    {
      "epoch": 1.7196333987978338,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 7223
    },
    {
      "epoch": 1.719871451526513,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 7224
    },
    {
      "epoch": 1.7201095042551926,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3725,
      "step": 7225
    },
    {
      "epoch": 1.7203475569838718,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 7226
    },
    {
      "epoch": 1.7205856097125514,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 7227
    },
    {
      "epoch": 1.7208236624412308,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 7228
    },
    {
      "epoch": 1.7210617151699101,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.455,
      "step": 7229
    },
    {
      "epoch": 1.7212997678985895,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5643,
      "step": 7230
    },
    {
      "epoch": 1.721537820627269,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 7231
    },
    {
      "epoch": 1.7217758733559485,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3446,
      "step": 7232
    },
    {
      "epoch": 1.7220139260846277,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 7233
    },
    {
      "epoch": 1.7222519788133073,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 7234
    },
    {
      "epoch": 1.7224900315419864,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 7235
    },
    {
      "epoch": 1.722728084270666,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 7236
    },
    {
      "epoch": 1.7229661369993452,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.552,
      "step": 7237
    },
    {
      "epoch": 1.7232041897280248,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 7238
    },
    {
      "epoch": 1.7234422424567042,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4448,
      "step": 7239
    },
    {
      "epoch": 1.7236802951853836,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3614,
      "step": 7240
    },
    {
      "epoch": 1.723918347914063,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 7241
    },
    {
      "epoch": 1.7241564006427423,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 7242
    },
    {
      "epoch": 1.7243944533714217,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 7243
    },
    {
      "epoch": 1.724632506100101,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 7244
    },
    {
      "epoch": 1.7248705588287807,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 7245
    },
    {
      "epoch": 1.7251086115574599,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4972,
      "step": 7246
    },
    {
      "epoch": 1.7253466642861395,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 7247
    },
    {
      "epoch": 1.7255847170148186,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 7248
    },
    {
      "epoch": 1.7258227697434982,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 7249
    },
    {
      "epoch": 1.7260608224721776,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 7250
    },
    {
      "epoch": 1.726298875200857,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 7251
    },
    {
      "epoch": 1.7265369279295364,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 7252
    },
    {
      "epoch": 1.7267749806582158,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3776,
      "step": 7253
    },
    {
      "epoch": 1.7270130333868952,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 7254
    },
    {
      "epoch": 1.7272510861155745,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 7255
    },
    {
      "epoch": 1.7274891388442541,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.458,
      "step": 7256
    },
    {
      "epoch": 1.7277271915729333,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.288,
      "step": 7257
    },
    {
      "epoch": 1.727965244301613,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4559,
      "step": 7258
    },
    {
      "epoch": 1.728203297030292,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 7259
    },
    {
      "epoch": 1.7284413497589717,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 7260
    },
    {
      "epoch": 1.728679402487651,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 7261
    },
    {
      "epoch": 1.7289174552163304,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2891,
      "step": 7262
    },
    {
      "epoch": 1.7291555079450098,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 7263
    },
    {
      "epoch": 1.7293935606736892,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 7264
    },
    {
      "epoch": 1.7296316134023686,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3936,
      "step": 7265
    },
    {
      "epoch": 1.729869666131048,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 7266
    },
    {
      "epoch": 1.7301077188597276,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 7267
    },
    {
      "epoch": 1.7303457715884067,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 7268
    },
    {
      "epoch": 1.7305838243170863,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 7269
    },
    {
      "epoch": 1.7308218770457655,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 7270
    },
    {
      "epoch": 1.7310599297744451,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 7271
    },
    {
      "epoch": 1.7312979825031243,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 7272
    },
    {
      "epoch": 1.7315360352318039,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 7273
    },
    {
      "epoch": 1.7317740879604833,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 7274
    },
    {
      "epoch": 1.7320121406891626,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 7275
    },
    {
      "epoch": 1.732250193417842,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 7276
    },
    {
      "epoch": 1.7324882461465214,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 7277
    },
    {
      "epoch": 1.732726298875201,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 7278
    },
    {
      "epoch": 1.7329643516038802,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 7279
    },
    {
      "epoch": 1.7332024043325598,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 7280
    },
    {
      "epoch": 1.733440457061239,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 7281
    },
    {
      "epoch": 1.7336785097899186,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 7282
    },
    {
      "epoch": 1.7339165625185977,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 7283
    },
    {
      "epoch": 1.7341546152472773,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3984,
      "step": 7284
    },
    {
      "epoch": 1.7343926679759567,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.5189,
      "step": 7285
    },
    {
      "epoch": 1.734630720704636,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 7286
    },
    {
      "epoch": 1.7348687734333155,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 7287
    },
    {
      "epoch": 1.7351068261619949,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4211,
      "step": 7288
    },
    {
      "epoch": 1.7353448788906742,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5756,
      "step": 7289
    },
    {
      "epoch": 1.7355829316193536,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 7290
    },
    {
      "epoch": 1.7358209843480332,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3163,
      "step": 7291
    },
    {
      "epoch": 1.7360590370767124,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 7292
    },
    {
      "epoch": 1.736297089805392,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 7293
    },
    {
      "epoch": 1.7365351425340712,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 7294
    },
    {
      "epoch": 1.7367731952627508,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 7295
    },
    {
      "epoch": 1.7370112479914301,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 7296
    },
    {
      "epoch": 1.7372493007201095,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3327,
      "step": 7297
    },
    {
      "epoch": 1.737487353448789,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 7298
    },
    {
      "epoch": 1.7377254061774683,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 7299
    },
    {
      "epoch": 1.7379634589061477,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 7300
    },
    {
      "epoch": 1.7379634589061477,
      "eval_runtime": 0.5242,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 7300
    },
    {
      "epoch": 1.738201511634827,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 7301
    },
    {
      "epoch": 1.7384395643635067,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 7302
    },
    {
      "epoch": 1.7386776170921858,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 7303
    },
    {
      "epoch": 1.7389156698208654,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 7304
    },
    {
      "epoch": 1.7391537225495446,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 7305
    },
    {
      "epoch": 1.7393917752782242,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 7306
    },
    {
      "epoch": 1.7396298280069036,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.3643,
      "step": 7307
    },
    {
      "epoch": 1.739867880735583,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4252,
      "step": 7308
    },
    {
      "epoch": 1.7401059334642623,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 7309
    },
    {
      "epoch": 1.7403439861929417,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 7310
    },
    {
      "epoch": 1.740582038921621,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 7311
    },
    {
      "epoch": 1.7408200916503005,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 7312
    },
    {
      "epoch": 1.74105814437898,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 7313
    },
    {
      "epoch": 1.7412961971076593,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 7314
    },
    {
      "epoch": 1.7415342498363389,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 7315
    },
    {
      "epoch": 1.741772302565018,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3908,
      "step": 7316
    },
    {
      "epoch": 1.7420103552936976,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 7317
    },
    {
      "epoch": 1.7422484080223768,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 7318
    },
    {
      "epoch": 1.7424864607510564,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3396,
      "step": 7319
    },
    {
      "epoch": 1.7427245134797358,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 7320
    },
    {
      "epoch": 1.7429625662084152,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5978,
      "step": 7321
    },
    {
      "epoch": 1.7432006189370945,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 7322
    },
    {
      "epoch": 1.743438671665774,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7905,
      "step": 7323
    },
    {
      "epoch": 1.7436767243944535,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4281,
      "step": 7324
    },
    {
      "epoch": 1.7439147771231327,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 7325
    },
    {
      "epoch": 1.7441528298518123,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4001,
      "step": 7326
    },
    {
      "epoch": 1.7443908825804915,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 7327
    },
    {
      "epoch": 1.744628935309171,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 7328
    },
    {
      "epoch": 1.7448669880378502,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 7329
    },
    {
      "epoch": 1.7451050407665298,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 7330
    },
    {
      "epoch": 1.7453430934952092,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 7331
    },
    {
      "epoch": 1.7455811462238886,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 7332
    },
    {
      "epoch": 1.745819198952568,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 7333
    },
    {
      "epoch": 1.7460572516812474,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4155,
      "step": 7334
    },
    {
      "epoch": 1.7462953044099268,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 7335
    },
    {
      "epoch": 1.7465333571386061,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 7336
    },
    {
      "epoch": 1.7467714098672857,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 7337
    },
    {
      "epoch": 1.747009462595965,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 7338
    },
    {
      "epoch": 1.7472475153246445,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 7339
    },
    {
      "epoch": 1.7474855680533237,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 7340
    },
    {
      "epoch": 1.7477236207820033,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 7341
    },
    {
      "epoch": 1.7479616735106827,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 7342
    },
    {
      "epoch": 1.748199726239362,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 7343
    },
    {
      "epoch": 1.7484377789680414,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 7344
    },
    {
      "epoch": 1.7486758316967208,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 7345
    },
    {
      "epoch": 1.7489138844254002,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 7346
    },
    {
      "epoch": 1.7491519371540796,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 7347
    },
    {
      "epoch": 1.7493899898827592,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 7348
    },
    {
      "epoch": 1.7496280426114383,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 7349
    },
    {
      "epoch": 1.749866095340118,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4482,
      "step": 7350
    },
    {
      "epoch": 1.750104148068797,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.5528,
      "step": 7351
    },
    {
      "epoch": 1.7503422007974767,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3127,
      "step": 7352
    },
    {
      "epoch": 1.750580253526156,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3896,
      "step": 7353
    },
    {
      "epoch": 1.7508183062548355,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 7354
    },
    {
      "epoch": 1.7510563589835149,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.436,
      "step": 7355
    },
    {
      "epoch": 1.7512944117121942,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 7356
    },
    {
      "epoch": 1.7515324644408736,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 7357
    },
    {
      "epoch": 1.751770517169553,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 7358
    },
    {
      "epoch": 1.7520085698982326,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 7359
    },
    {
      "epoch": 1.7522466226269118,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 7360
    },
    {
      "epoch": 1.7524846753555914,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3899,
      "step": 7361
    },
    {
      "epoch": 1.7527227280842705,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 7362
    },
    {
      "epoch": 1.7529607808129501,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 7363
    },
    {
      "epoch": 1.7531988335416293,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 7364
    },
    {
      "epoch": 1.753436886270309,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 7365
    },
    {
      "epoch": 1.7536749389989883,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 7366
    },
    {
      "epoch": 1.7539129917276677,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.5432,
      "step": 7367
    },
    {
      "epoch": 1.754151044456347,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 7368
    },
    {
      "epoch": 1.7543890971850264,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 7369
    },
    {
      "epoch": 1.754627149913706,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 7370
    },
    {
      "epoch": 1.7548652026423852,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 7371
    },
    {
      "epoch": 1.7551032553710648,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.3375,
      "step": 7372
    },
    {
      "epoch": 1.755341308099744,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 7373
    },
    {
      "epoch": 1.7555793608284236,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 7374
    },
    {
      "epoch": 1.7558174135571027,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 7375
    },
    {
      "epoch": 1.7560554662857824,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 7376
    },
    {
      "epoch": 1.7562935190144617,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 7377
    },
    {
      "epoch": 1.7565315717431411,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 7378
    },
    {
      "epoch": 1.7567696244718205,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 7379
    },
    {
      "epoch": 1.7570076772004999,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 7380
    },
    {
      "epoch": 1.7572457299291793,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 7381
    },
    {
      "epoch": 1.7574837826578587,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 7382
    },
    {
      "epoch": 1.7577218353865383,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3326,
      "step": 7383
    },
    {
      "epoch": 1.7579598881152174,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 7384
    },
    {
      "epoch": 1.758197940843897,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 7385
    },
    {
      "epoch": 1.7584359935725762,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 7386
    },
    {
      "epoch": 1.7586740463012558,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 7387
    },
    {
      "epoch": 1.7589120990299352,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 7388
    },
    {
      "epoch": 1.7591501517586146,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4088,
      "step": 7389
    },
    {
      "epoch": 1.759388204487294,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 7390
    },
    {
      "epoch": 1.7596262572159733,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 7391
    },
    {
      "epoch": 1.7598643099446527,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 7392
    },
    {
      "epoch": 1.760102362673332,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 7393
    },
    {
      "epoch": 1.7603404154020117,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 7394
    },
    {
      "epoch": 1.7605784681306909,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 7395
    },
    {
      "epoch": 1.7608165208593705,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 7396
    },
    {
      "epoch": 1.7610545735880496,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4308,
      "step": 7397
    },
    {
      "epoch": 1.7612926263167292,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 7398
    },
    {
      "epoch": 1.7615306790454086,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2428,
      "step": 7399
    },
    {
      "epoch": 1.761768731774088,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 7400
    },
    {
      "epoch": 1.761768731774088,
      "eval_runtime": 0.5226,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 7400
    },
    {
      "epoch": 1.7620067845027674,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3789,
      "step": 7401
    },
    {
      "epoch": 1.7622448372314468,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5225,
      "step": 7402
    },
    {
      "epoch": 1.7624828899601261,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 7403
    },
    {
      "epoch": 1.7627209426888055,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 7404
    },
    {
      "epoch": 1.7629589954174851,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 7405
    },
    {
      "epoch": 1.7631970481461643,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.443,
      "step": 7406
    },
    {
      "epoch": 1.763435100874844,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 7407
    },
    {
      "epoch": 1.763673153603523,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4339,
      "step": 7408
    },
    {
      "epoch": 1.7639112063322027,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 7409
    },
    {
      "epoch": 1.7641492590608818,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 7410
    },
    {
      "epoch": 1.7643873117895614,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4226,
      "step": 7411
    },
    {
      "epoch": 1.7646253645182408,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 7412
    },
    {
      "epoch": 1.7648634172469202,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4717,
      "step": 7413
    },
    {
      "epoch": 1.7651014699755996,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 7414
    },
    {
      "epoch": 1.765339522704279,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 7415
    },
    {
      "epoch": 1.7655775754329583,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 7416
    },
    {
      "epoch": 1.7658156281616377,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4097,
      "step": 7417
    },
    {
      "epoch": 1.7660536808903173,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 7418
    },
    {
      "epoch": 1.7662917336189965,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 7419
    },
    {
      "epoch": 1.766529786347676,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2999,
      "step": 7420
    },
    {
      "epoch": 1.7667678390763553,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 7421
    },
    {
      "epoch": 1.7670058918050349,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 7422
    },
    {
      "epoch": 1.7672439445337143,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 7423
    },
    {
      "epoch": 1.7674819972623936,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 7424
    },
    {
      "epoch": 1.767720049991073,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 7425
    },
    {
      "epoch": 1.7679581027197524,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 7426
    },
    {
      "epoch": 1.7681961554484318,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 7427
    },
    {
      "epoch": 1.7684342081771112,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 7428
    },
    {
      "epoch": 1.7686722609057908,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 7429
    },
    {
      "epoch": 1.76891031363447,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2748,
      "step": 7430
    },
    {
      "epoch": 1.7691483663631495,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 7431
    },
    {
      "epoch": 1.7693864190918287,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 7432
    },
    {
      "epoch": 1.7696244718205083,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 7433
    },
    {
      "epoch": 1.7698625245491877,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 7434
    },
    {
      "epoch": 1.770100577277867,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 7435
    },
    {
      "epoch": 1.7703386300065465,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 7436
    },
    {
      "epoch": 1.7705766827352258,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 7437
    },
    {
      "epoch": 1.7708147354639052,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 7438
    },
    {
      "epoch": 1.7710527881925846,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 7439
    },
    {
      "epoch": 1.7712908409212642,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 7440
    },
    {
      "epoch": 1.7715288936499434,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 7441
    },
    {
      "epoch": 1.771766946378623,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 7442
    },
    {
      "epoch": 1.7720049991073021,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 7443
    },
    {
      "epoch": 1.7722430518359817,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 7444
    },
    {
      "epoch": 1.7724811045646611,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 7445
    },
    {
      "epoch": 1.7727191572933405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 7446
    },
    {
      "epoch": 1.77295721002202,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 7447
    },
    {
      "epoch": 1.7731952627506993,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 7448
    },
    {
      "epoch": 1.7734333154793787,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 7449
    },
    {
      "epoch": 1.773671368208058,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 7450
    },
    {
      "epoch": 1.7739094209367376,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.358,
      "step": 7451
    },
    {
      "epoch": 1.7741474736654168,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.6724,
      "step": 7452
    },
    {
      "epoch": 1.7743855263940964,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 7453
    },
    {
      "epoch": 1.7746235791227756,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 7454
    },
    {
      "epoch": 1.7748616318514552,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 7455
    },
    {
      "epoch": 1.7750996845801343,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 7456
    },
    {
      "epoch": 1.775337737308814,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 7457
    },
    {
      "epoch": 1.7755757900374933,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 7458
    },
    {
      "epoch": 1.7758138427661727,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 7459
    },
    {
      "epoch": 1.776051895494852,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 7460
    },
    {
      "epoch": 1.7762899482235315,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 7461
    },
    {
      "epoch": 1.7765280009522109,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 7462
    },
    {
      "epoch": 1.7767660536808902,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3447,
      "step": 7463
    },
    {
      "epoch": 1.7770041064095699,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 7464
    },
    {
      "epoch": 1.777242159138249,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 7465
    },
    {
      "epoch": 1.7774802118669286,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3424,
      "step": 7466
    },
    {
      "epoch": 1.7777182645956078,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 7467
    },
    {
      "epoch": 1.7779563173242874,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 7468
    },
    {
      "epoch": 1.7781943700529668,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 7469
    },
    {
      "epoch": 1.7784324227816461,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 7470
    },
    {
      "epoch": 1.7786704755103255,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 7471
    },
    {
      "epoch": 1.778908528239005,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 7472
    },
    {
      "epoch": 1.7791465809676843,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 7473
    },
    {
      "epoch": 1.7793846336963637,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.371,
      "step": 7474
    },
    {
      "epoch": 1.7796226864250433,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 7475
    },
    {
      "epoch": 1.7798607391537224,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 7476
    },
    {
      "epoch": 1.780098791882402,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 7477
    },
    {
      "epoch": 1.7803368446110812,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 7478
    },
    {
      "epoch": 1.7805748973397608,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4911,
      "step": 7479
    },
    {
      "epoch": 1.7808129500684402,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.5028,
      "step": 7480
    },
    {
      "epoch": 1.7810510027971196,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 7481
    },
    {
      "epoch": 1.781289055525799,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 7482
    },
    {
      "epoch": 1.7815271082544784,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 7483
    },
    {
      "epoch": 1.7817651609831577,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 7484
    },
    {
      "epoch": 1.7820032137118371,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 7485
    },
    {
      "epoch": 1.7822412664405167,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 7486
    },
    {
      "epoch": 1.7824793191691959,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 7487
    },
    {
      "epoch": 1.7827173718978755,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4177,
      "step": 7488
    },
    {
      "epoch": 1.7829554246265547,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 7489
    },
    {
      "epoch": 1.7831934773552343,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 7490
    },
    {
      "epoch": 1.7834315300839136,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 7491
    },
    {
      "epoch": 1.783669582812593,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 7492
    },
    {
      "epoch": 1.7839076355412724,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2778,
      "step": 7493
    },
    {
      "epoch": 1.7841456882699518,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 7494
    },
    {
      "epoch": 1.7843837409986312,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 7495
    },
    {
      "epoch": 1.7846217937273106,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 7496
    },
    {
      "epoch": 1.7848598464559902,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 7497
    },
    {
      "epoch": 1.7850978991846693,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 7498
    },
    {
      "epoch": 1.785335951913349,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 7499
    },
    {
      "epoch": 1.785574004642028,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 7500
    },
    {
      "epoch": 1.785574004642028,
      "eval_runtime": 0.5336,
      "eval_samples_per_second": 1.874,
      "eval_steps_per_second": 1.874,
      "step": 7500
    },
    {
      "epoch": 1.7858120573707077,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 7501
    },
    {
      "epoch": 1.7860501100993869,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 7502
    },
    {
      "epoch": 1.7862881628280665,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 7503
    },
    {
      "epoch": 1.7865262155567458,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 7504
    },
    {
      "epoch": 1.7867642682854252,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 7505
    },
    {
      "epoch": 1.7870023210141046,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 7506
    },
    {
      "epoch": 1.787240373742784,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3585,
      "step": 7507
    },
    {
      "epoch": 1.7874784264714634,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 7508
    },
    {
      "epoch": 1.7877164792001428,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4171,
      "step": 7509
    },
    {
      "epoch": 1.7879545319288224,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 7510
    },
    {
      "epoch": 1.7881925846575015,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 7511
    },
    {
      "epoch": 1.7884306373861811,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4959,
      "step": 7512
    },
    {
      "epoch": 1.7886686901148603,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 7513
    },
    {
      "epoch": 1.78890674284354,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 7514
    },
    {
      "epoch": 1.7891447955722193,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 7515
    },
    {
      "epoch": 1.7893828483008987,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 7516
    },
    {
      "epoch": 1.789620901029578,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 7517
    },
    {
      "epoch": 1.7898589537582574,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 7518
    },
    {
      "epoch": 1.7900970064869368,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3297,
      "step": 7519
    },
    {
      "epoch": 1.7903350592156162,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 7520
    },
    {
      "epoch": 1.7905731119442958,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 7521
    },
    {
      "epoch": 1.790811164672975,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 7522
    },
    {
      "epoch": 1.7910492174016546,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3256,
      "step": 7523
    },
    {
      "epoch": 1.7912872701303337,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 7524
    },
    {
      "epoch": 1.7915253228590133,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 7525
    },
    {
      "epoch": 1.7917633755876927,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 7526
    },
    {
      "epoch": 1.792001428316372,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 7527
    },
    {
      "epoch": 1.7922394810450515,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.332,
      "step": 7528
    },
    {
      "epoch": 1.7924775337737309,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 7529
    },
    {
      "epoch": 1.7927155865024103,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 7530
    },
    {
      "epoch": 1.7929536392310896,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 7531
    },
    {
      "epoch": 1.7931916919597692,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.273,
      "step": 7532
    },
    {
      "epoch": 1.7934297446884484,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 7533
    },
    {
      "epoch": 1.793667797417128,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 7534
    },
    {
      "epoch": 1.7939058501458072,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 7535
    },
    {
      "epoch": 1.7941439028744868,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 7536
    },
    {
      "epoch": 1.7943819556031662,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 7537
    },
    {
      "epoch": 1.7946200083318455,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 7538
    },
    {
      "epoch": 1.794858061060525,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 7539
    },
    {
      "epoch": 1.7950961137892043,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 7540
    },
    {
      "epoch": 1.7953341665178837,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 7541
    },
    {
      "epoch": 1.795572219246563,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 7542
    },
    {
      "epoch": 1.7958102719752427,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 7543
    },
    {
      "epoch": 1.7960483247039218,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3774,
      "step": 7544
    },
    {
      "epoch": 1.7962863774326014,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 7545
    },
    {
      "epoch": 1.7965244301612806,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 7546
    },
    {
      "epoch": 1.7967624828899602,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4637,
      "step": 7547
    },
    {
      "epoch": 1.7970005356186394,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 7548
    },
    {
      "epoch": 1.797238588347319,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 7549
    },
    {
      "epoch": 1.7974766410759984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 7550
    },
    {
      "epoch": 1.7977146938046777,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 7551
    },
    {
      "epoch": 1.7979527465333571,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 7552
    },
    {
      "epoch": 1.7981907992620365,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 7553
    },
    {
      "epoch": 1.798428851990716,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 7554
    },
    {
      "epoch": 1.7986669047193953,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 7555
    },
    {
      "epoch": 1.7989049574480749,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 7556
    },
    {
      "epoch": 1.799143010176754,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 7557
    },
    {
      "epoch": 1.7993810629054336,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 7558
    },
    {
      "epoch": 1.7996191156341128,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4021,
      "step": 7559
    },
    {
      "epoch": 1.7998571683627924,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 7560
    },
    {
      "epoch": 1.8000952210914718,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3806,
      "step": 7561
    },
    {
      "epoch": 1.8003332738201512,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 7562
    },
    {
      "epoch": 1.8005713265488306,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 7563
    },
    {
      "epoch": 1.80080937927751,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 7564
    },
    {
      "epoch": 1.8010474320061893,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5479,
      "step": 7565
    },
    {
      "epoch": 1.8012854847348687,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 7566
    },
    {
      "epoch": 1.8015235374635483,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 7567
    },
    {
      "epoch": 1.8017615901922275,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 7568
    },
    {
      "epoch": 1.801999642920907,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4546,
      "step": 7569
    },
    {
      "epoch": 1.8022376956495862,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 7570
    },
    {
      "epoch": 1.8024757483782659,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 7571
    },
    {
      "epoch": 1.8027138011069452,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 7572
    },
    {
      "epoch": 1.8029518538356246,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 7573
    },
    {
      "epoch": 1.803189906564304,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 7574
    },
    {
      "epoch": 1.8034279592929834,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 7575
    },
    {
      "epoch": 1.8036660120216628,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 7576
    },
    {
      "epoch": 1.8039040647503422,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 7577
    },
    {
      "epoch": 1.8041421174790218,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 7578
    },
    {
      "epoch": 1.804380170207701,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 7579
    },
    {
      "epoch": 1.8046182229363805,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5318,
      "step": 7580
    },
    {
      "epoch": 1.8048562756650597,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 7581
    },
    {
      "epoch": 1.8050943283937393,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.5097,
      "step": 7582
    },
    {
      "epoch": 1.8053323811224185,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 7583
    },
    {
      "epoch": 1.805570433851098,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 7584
    },
    {
      "epoch": 1.8058084865797774,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3258,
      "step": 7585
    },
    {
      "epoch": 1.8060465393084568,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 7586
    },
    {
      "epoch": 1.8062845920371362,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.448,
      "step": 7587
    },
    {
      "epoch": 1.8065226447658156,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 7588
    },
    {
      "epoch": 1.8067606974944952,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4,
      "step": 7589
    },
    {
      "epoch": 1.8069987502231744,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 7590
    },
    {
      "epoch": 1.807236802951854,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 7591
    },
    {
      "epoch": 1.8074748556805331,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 7592
    },
    {
      "epoch": 1.8077129084092127,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 7593
    },
    {
      "epoch": 1.8079509611378919,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 7594
    },
    {
      "epoch": 1.8081890138665715,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 7595
    },
    {
      "epoch": 1.8084270665952509,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 7596
    },
    {
      "epoch": 1.8086651193239303,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 7597
    },
    {
      "epoch": 1.8089031720526096,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 7598
    },
    {
      "epoch": 1.809141224781289,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 7599
    },
    {
      "epoch": 1.8093792775099684,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 7600
    },
    {
      "epoch": 1.8093792775099684,
      "eval_runtime": 0.5257,
      "eval_samples_per_second": 1.902,
      "eval_steps_per_second": 1.902,
      "step": 7600
    },
    {
      "epoch": 1.8096173302386478,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 7601
    },
    {
      "epoch": 1.8098553829673274,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4207,
      "step": 7602
    },
    {
      "epoch": 1.8100934356960066,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3204,
      "step": 7603
    },
    {
      "epoch": 1.8103314884246862,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 7604
    },
    {
      "epoch": 1.8105695411533653,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 7605
    },
    {
      "epoch": 1.810807593882045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 7606
    },
    {
      "epoch": 1.8110456466107243,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 7607
    },
    {
      "epoch": 1.8112836993394037,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 7608
    },
    {
      "epoch": 1.811521752068083,
      "grad_norm": 8.3125,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 7609
    },
    {
      "epoch": 1.8117598047967625,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 7610
    },
    {
      "epoch": 1.8119978575254418,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 7611
    },
    {
      "epoch": 1.8122359102541212,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 7612
    },
    {
      "epoch": 1.8124739629828008,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 7613
    },
    {
      "epoch": 1.81271201571148,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 7614
    },
    {
      "epoch": 1.8129500684401596,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 7615
    },
    {
      "epoch": 1.8131881211688388,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 7616
    },
    {
      "epoch": 1.8134261738975184,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 7617
    },
    {
      "epoch": 1.8136642266261978,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3777,
      "step": 7618
    },
    {
      "epoch": 1.8139022793548771,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4231,
      "step": 7619
    },
    {
      "epoch": 1.8141403320835565,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 7620
    },
    {
      "epoch": 1.814378384812236,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 7621
    },
    {
      "epoch": 1.8146164375409153,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 7622
    },
    {
      "epoch": 1.8148544902695947,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 7623
    },
    {
      "epoch": 1.8150925429982743,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 7624
    },
    {
      "epoch": 1.8153305957269534,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 7625
    },
    {
      "epoch": 1.815568648455633,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 7626
    },
    {
      "epoch": 1.8158067011843122,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 7627
    },
    {
      "epoch": 1.8160447539129918,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 7628
    },
    {
      "epoch": 1.816282806641671,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 7629
    },
    {
      "epoch": 1.8165208593703506,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 7630
    },
    {
      "epoch": 1.81675891209903,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 7631
    },
    {
      "epoch": 1.8169969648277093,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 7632
    },
    {
      "epoch": 1.8172350175563887,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 7633
    },
    {
      "epoch": 1.817473070285068,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 7634
    },
    {
      "epoch": 1.8177111230137477,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 7635
    },
    {
      "epoch": 1.8179491757424269,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 7636
    },
    {
      "epoch": 1.8181872284711065,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4426,
      "step": 7637
    },
    {
      "epoch": 1.8184252811997856,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.357,
      "step": 7638
    },
    {
      "epoch": 1.8186633339284652,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 7639
    },
    {
      "epoch": 1.8189013866571444,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 7640
    },
    {
      "epoch": 1.819139439385824,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 7641
    },
    {
      "epoch": 1.8193774921145034,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 7642
    },
    {
      "epoch": 1.8196155448431828,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.6601,
      "step": 7643
    },
    {
      "epoch": 1.8198535975718622,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 7644
    },
    {
      "epoch": 1.8200916503005415,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 7645
    },
    {
      "epoch": 1.820329703029221,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 7646
    },
    {
      "epoch": 1.8205677557579003,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 7647
    },
    {
      "epoch": 1.82080580848658,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4345,
      "step": 7648
    },
    {
      "epoch": 1.821043861215259,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 7649
    },
    {
      "epoch": 1.8212819139439387,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 7650
    },
    {
      "epoch": 1.8215199666726178,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 7651
    },
    {
      "epoch": 1.8217580194012974,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5387,
      "step": 7652
    },
    {
      "epoch": 1.8219960721299768,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 7653
    },
    {
      "epoch": 1.8222341248586562,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 7654
    },
    {
      "epoch": 1.8224721775873356,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 7655
    },
    {
      "epoch": 1.822710230316015,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4417,
      "step": 7656
    },
    {
      "epoch": 1.8229482830446944,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2995,
      "step": 7657
    },
    {
      "epoch": 1.8231863357733737,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 7658
    },
    {
      "epoch": 1.8234243885020534,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4922,
      "step": 7659
    },
    {
      "epoch": 1.8236624412307325,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 7660
    },
    {
      "epoch": 1.8239004939594121,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 7661
    },
    {
      "epoch": 1.8241385466880913,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 7662
    },
    {
      "epoch": 1.8243765994167709,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 7663
    },
    {
      "epoch": 1.8246146521454503,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.649,
      "step": 7664
    },
    {
      "epoch": 1.8248527048741296,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 7665
    },
    {
      "epoch": 1.825090757602809,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 7666
    },
    {
      "epoch": 1.8253288103314884,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 7667
    },
    {
      "epoch": 1.8255668630601678,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 7668
    },
    {
      "epoch": 1.8258049157888472,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 7669
    },
    {
      "epoch": 1.8260429685175268,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 7670
    },
    {
      "epoch": 1.826281021246206,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.7315,
      "step": 7671
    },
    {
      "epoch": 1.8265190739748856,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 7672
    },
    {
      "epoch": 1.8267571267035647,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 7673
    },
    {
      "epoch": 1.8269951794322443,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 7674
    },
    {
      "epoch": 1.8272332321609235,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 7675
    },
    {
      "epoch": 1.827471284889603,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4878,
      "step": 7676
    },
    {
      "epoch": 1.8277093376182825,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 7677
    },
    {
      "epoch": 1.8279473903469619,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 7678
    },
    {
      "epoch": 1.8281854430756412,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4268,
      "step": 7679
    },
    {
      "epoch": 1.8284234958043206,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 7680
    },
    {
      "epoch": 1.8286615485330002,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 7681
    },
    {
      "epoch": 1.8288996012616794,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 7682
    },
    {
      "epoch": 1.829137653990359,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 7683
    },
    {
      "epoch": 1.8293757067190382,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 7684
    },
    {
      "epoch": 1.8296137594477178,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 7685
    },
    {
      "epoch": 1.829851812176397,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 7686
    },
    {
      "epoch": 1.8300898649050765,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 7687
    },
    {
      "epoch": 1.830327917633756,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 7688
    },
    {
      "epoch": 1.8305659703624353,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 7689
    },
    {
      "epoch": 1.8308040230911147,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3579,
      "step": 7690
    },
    {
      "epoch": 1.831042075819794,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 7691
    },
    {
      "epoch": 1.8312801285484734,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 7692
    },
    {
      "epoch": 1.8315181812771528,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 7693
    },
    {
      "epoch": 1.8317562340058324,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 7694
    },
    {
      "epoch": 1.8319942867345116,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2549,
      "step": 7695
    },
    {
      "epoch": 1.8322323394631912,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 7696
    },
    {
      "epoch": 1.8324703921918704,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 7697
    },
    {
      "epoch": 1.83270844492055,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 7698
    },
    {
      "epoch": 1.8329464976492293,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 7699
    },
    {
      "epoch": 1.8331845503779087,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 7700
    },
    {
      "epoch": 1.8331845503779087,
      "eval_runtime": 0.5357,
      "eval_samples_per_second": 1.867,
      "eval_steps_per_second": 1.867,
      "step": 7700
    },
    {
      "epoch": 1.8334226031065881,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4819,
      "step": 7701
    },
    {
      "epoch": 1.8336606558352675,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 7702
    },
    {
      "epoch": 1.8338987085639469,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 7703
    },
    {
      "epoch": 1.8341367612926263,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 7704
    },
    {
      "epoch": 1.8343748140213059,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 7705
    },
    {
      "epoch": 1.834612866749985,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 7706
    },
    {
      "epoch": 1.8348509194786646,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 7707
    },
    {
      "epoch": 1.8350889722073438,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3286,
      "step": 7708
    },
    {
      "epoch": 1.8353270249360234,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 7709
    },
    {
      "epoch": 1.8355650776647028,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 7710
    },
    {
      "epoch": 1.8358031303933822,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 7711
    },
    {
      "epoch": 1.8360411831220615,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 7712
    },
    {
      "epoch": 1.836279235850741,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2235,
      "step": 7713
    },
    {
      "epoch": 1.8365172885794203,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 7714
    },
    {
      "epoch": 1.8367553413080997,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7696,
      "step": 7715
    },
    {
      "epoch": 1.8369933940367793,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 7716
    },
    {
      "epoch": 1.8372314467654585,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.359,
      "step": 7717
    },
    {
      "epoch": 1.837469499494138,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 7718
    },
    {
      "epoch": 1.8377075522228172,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8765,
      "step": 7719
    },
    {
      "epoch": 1.8379456049514968,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 7720
    },
    {
      "epoch": 1.838183657680176,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 7721
    },
    {
      "epoch": 1.8384217104088556,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 7722
    },
    {
      "epoch": 1.838659763137535,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.5192,
      "step": 7723
    },
    {
      "epoch": 1.8388978158662144,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 7724
    },
    {
      "epoch": 1.8391358685948938,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3395,
      "step": 7725
    },
    {
      "epoch": 1.8393739213235731,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 7726
    },
    {
      "epoch": 1.8396119740522525,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 7727
    },
    {
      "epoch": 1.839850026780932,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 7728
    },
    {
      "epoch": 1.8400880795096115,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 7729
    },
    {
      "epoch": 1.8403261322382907,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 7730
    },
    {
      "epoch": 1.8405641849669703,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 7731
    },
    {
      "epoch": 1.8408022376956494,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 7732
    },
    {
      "epoch": 1.841040290424329,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 7733
    },
    {
      "epoch": 1.8412783431530084,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 7734
    },
    {
      "epoch": 1.8415163958816878,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 7735
    },
    {
      "epoch": 1.8417544486103672,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 7736
    },
    {
      "epoch": 1.8419925013390466,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 7737
    },
    {
      "epoch": 1.842230554067726,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 7738
    },
    {
      "epoch": 1.8424686067964053,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 7739
    },
    {
      "epoch": 1.842706659525085,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.616,
      "step": 7740
    },
    {
      "epoch": 1.842944712253764,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 7741
    },
    {
      "epoch": 1.8431827649824437,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 7742
    },
    {
      "epoch": 1.8434208177111229,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 7743
    },
    {
      "epoch": 1.8436588704398025,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 7744
    },
    {
      "epoch": 1.8438969231684819,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 7745
    },
    {
      "epoch": 1.8441349758971612,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 7746
    },
    {
      "epoch": 1.8443730286258406,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 7747
    },
    {
      "epoch": 1.84461108135452,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 7748
    },
    {
      "epoch": 1.8448491340831994,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2846,
      "step": 7749
    },
    {
      "epoch": 1.8450871868118788,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 7750
    },
    {
      "epoch": 1.8453252395405584,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 7751
    },
    {
      "epoch": 1.8455632922692375,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 7752
    },
    {
      "epoch": 1.8458013449979171,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 7753
    },
    {
      "epoch": 1.8460393977265963,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 7754
    },
    {
      "epoch": 1.846277450455276,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 7755
    },
    {
      "epoch": 1.8465155031839553,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5332,
      "step": 7756
    },
    {
      "epoch": 1.8467535559126347,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3977,
      "step": 7757
    },
    {
      "epoch": 1.846991608641314,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 7758
    },
    {
      "epoch": 1.8472296613699934,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 7759
    },
    {
      "epoch": 1.8474677140986728,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 7760
    },
    {
      "epoch": 1.8477057668273522,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 7761
    },
    {
      "epoch": 1.8479438195560318,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 7762
    },
    {
      "epoch": 1.848181872284711,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 7763
    },
    {
      "epoch": 1.8484199250133906,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 7764
    },
    {
      "epoch": 1.8486579777420697,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 7765
    },
    {
      "epoch": 1.8488960304707494,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 7766
    },
    {
      "epoch": 1.8491340831994285,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 7767
    },
    {
      "epoch": 1.8493721359281081,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 7768
    },
    {
      "epoch": 1.8496101886567875,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 7769
    },
    {
      "epoch": 1.8498482413854669,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 7770
    },
    {
      "epoch": 1.8500862941141463,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 7771
    },
    {
      "epoch": 1.8503243468428257,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 7772
    },
    {
      "epoch": 1.850562399571505,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 7773
    },
    {
      "epoch": 1.8508004523001844,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 7774
    },
    {
      "epoch": 1.851038505028864,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 7775
    },
    {
      "epoch": 1.8512765577575432,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 7776
    },
    {
      "epoch": 1.8515146104862228,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 7777
    },
    {
      "epoch": 1.851752663214902,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3974,
      "step": 7778
    },
    {
      "epoch": 1.8519907159435816,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 7779
    },
    {
      "epoch": 1.852228768672261,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 7780
    },
    {
      "epoch": 1.8524668214009403,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 7781
    },
    {
      "epoch": 1.8527048741296197,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3359,
      "step": 7782
    },
    {
      "epoch": 1.852942926858299,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 7783
    },
    {
      "epoch": 1.8531809795869785,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 7784
    },
    {
      "epoch": 1.8534190323156579,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 7785
    },
    {
      "epoch": 1.8536570850443375,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 7786
    },
    {
      "epoch": 1.8538951377730166,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 7787
    },
    {
      "epoch": 1.8541331905016962,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 7788
    },
    {
      "epoch": 1.8543712432303754,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.376,
      "step": 7789
    },
    {
      "epoch": 1.854609295959055,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 7790
    },
    {
      "epoch": 1.8548473486877344,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 7791
    },
    {
      "epoch": 1.8550854014164138,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2757,
      "step": 7792
    },
    {
      "epoch": 1.8553234541450931,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 7793
    },
    {
      "epoch": 1.8555615068737725,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 7794
    },
    {
      "epoch": 1.855799559602452,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 7795
    },
    {
      "epoch": 1.8560376123311313,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 7796
    },
    {
      "epoch": 1.856275665059811,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 7797
    },
    {
      "epoch": 1.85651371778849,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 7798
    },
    {
      "epoch": 1.8567517705171697,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 7799
    },
    {
      "epoch": 1.8569898232458488,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 7800
    },
    {
      "epoch": 1.8569898232458488,
      "eval_runtime": 0.5227,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 7800
    },
    {
      "epoch": 1.8572278759745284,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 7801
    },
    {
      "epoch": 1.8574659287032078,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 7802
    },
    {
      "epoch": 1.8577039814318872,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 7803
    },
    {
      "epoch": 1.8579420341605666,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 7804
    },
    {
      "epoch": 1.858180086889246,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 7805
    },
    {
      "epoch": 1.8584181396179253,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 7806
    },
    {
      "epoch": 1.8586561923466047,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 7807
    },
    {
      "epoch": 1.8588942450752843,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 7808
    },
    {
      "epoch": 1.8591322978039635,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 7809
    },
    {
      "epoch": 1.859370350532643,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 7810
    },
    {
      "epoch": 1.8596084032613223,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2525,
      "step": 7811
    },
    {
      "epoch": 1.8598464559900019,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3245,
      "step": 7812
    },
    {
      "epoch": 1.860084508718681,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 7813
    },
    {
      "epoch": 1.8603225614473606,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 7814
    },
    {
      "epoch": 1.86056061417604,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 7815
    },
    {
      "epoch": 1.8607986669047194,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 7816
    },
    {
      "epoch": 1.8610367196333988,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 7817
    },
    {
      "epoch": 1.8612747723620782,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 7818
    },
    {
      "epoch": 1.8615128250907576,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 7819
    },
    {
      "epoch": 1.861750877819437,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 7820
    },
    {
      "epoch": 1.8619889305481165,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 7821
    },
    {
      "epoch": 1.8622269832767957,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 7822
    },
    {
      "epoch": 1.8624650360054753,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 7823
    },
    {
      "epoch": 1.8627030887341545,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.511,
      "step": 7824
    },
    {
      "epoch": 1.862941141462834,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4048,
      "step": 7825
    },
    {
      "epoch": 1.8631791941915135,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 7826
    },
    {
      "epoch": 1.8634172469201928,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 7827
    },
    {
      "epoch": 1.8636552996488722,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 7828
    },
    {
      "epoch": 1.8638933523775516,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 7829
    },
    {
      "epoch": 1.864131405106231,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 7830
    },
    {
      "epoch": 1.8643694578349104,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 7831
    },
    {
      "epoch": 1.86460751056359,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 7832
    },
    {
      "epoch": 1.8648455632922691,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3632,
      "step": 7833
    },
    {
      "epoch": 1.8650836160209487,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 7834
    },
    {
      "epoch": 1.865321668749628,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 7835
    },
    {
      "epoch": 1.8655597214783075,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 7836
    },
    {
      "epoch": 1.865797774206987,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3536,
      "step": 7837
    },
    {
      "epoch": 1.8660358269356663,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 7838
    },
    {
      "epoch": 1.8662738796643457,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 7839
    },
    {
      "epoch": 1.866511932393025,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 7840
    },
    {
      "epoch": 1.8667499851217044,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 7841
    },
    {
      "epoch": 1.8669880378503838,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 7842
    },
    {
      "epoch": 1.8672260905790634,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 7843
    },
    {
      "epoch": 1.8674641433077426,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 7844
    },
    {
      "epoch": 1.8677021960364222,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 7845
    },
    {
      "epoch": 1.8679402487651013,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 7846
    },
    {
      "epoch": 1.868178301493781,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4296,
      "step": 7847
    },
    {
      "epoch": 1.8684163542224603,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 7848
    },
    {
      "epoch": 1.8686544069511397,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4064,
      "step": 7849
    },
    {
      "epoch": 1.868892459679819,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 7850
    },
    {
      "epoch": 1.8691305124084985,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 7851
    },
    {
      "epoch": 1.8693685651371779,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 7852
    },
    {
      "epoch": 1.8696066178658572,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 7853
    },
    {
      "epoch": 1.8698446705945369,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 7854
    },
    {
      "epoch": 1.870082723323216,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 7855
    },
    {
      "epoch": 1.8703207760518956,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4221,
      "step": 7856
    },
    {
      "epoch": 1.8705588287805748,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 7857
    },
    {
      "epoch": 1.8707968815092544,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 7858
    },
    {
      "epoch": 1.8710349342379335,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4584,
      "step": 7859
    },
    {
      "epoch": 1.8712729869666131,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 7860
    },
    {
      "epoch": 1.8715110396952925,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 7861
    },
    {
      "epoch": 1.871749092423972,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 7862
    },
    {
      "epoch": 1.8719871451526513,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 7863
    },
    {
      "epoch": 1.8722251978813307,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 7864
    },
    {
      "epoch": 1.87246325061001,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 7865
    },
    {
      "epoch": 1.8727013033386894,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3227,
      "step": 7866
    },
    {
      "epoch": 1.872939356067369,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 7867
    },
    {
      "epoch": 1.8731774087960482,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 7868
    },
    {
      "epoch": 1.8734154615247278,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 7869
    },
    {
      "epoch": 1.873653514253407,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 7870
    },
    {
      "epoch": 1.8738915669820866,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 7871
    },
    {
      "epoch": 1.874129619710766,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3058,
      "step": 7872
    },
    {
      "epoch": 1.8743676724394454,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 7873
    },
    {
      "epoch": 1.8746057251681247,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 7874
    },
    {
      "epoch": 1.8748437778968041,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 7875
    },
    {
      "epoch": 1.8750818306254835,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 7876
    },
    {
      "epoch": 1.8753198833541629,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 7877
    },
    {
      "epoch": 1.8755579360828425,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 7878
    },
    {
      "epoch": 1.8757959888115217,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 7879
    },
    {
      "epoch": 1.8760340415402013,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 7880
    },
    {
      "epoch": 1.8762720942688804,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 7881
    },
    {
      "epoch": 1.87651014699756,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 7882
    },
    {
      "epoch": 1.8767481997262394,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 7883
    },
    {
      "epoch": 1.8769862524549188,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 7884
    },
    {
      "epoch": 1.8772243051835982,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.4277,
      "step": 7885
    },
    {
      "epoch": 1.8774623579122776,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 7886
    },
    {
      "epoch": 1.877700410640957,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3437,
      "step": 7887
    },
    {
      "epoch": 1.8779384633696363,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 7888
    },
    {
      "epoch": 1.878176516098316,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2816,
      "step": 7889
    },
    {
      "epoch": 1.878414568826995,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2454,
      "step": 7890
    },
    {
      "epoch": 1.8786526215556747,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.5311,
      "step": 7891
    },
    {
      "epoch": 1.8788906742843539,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4555,
      "step": 7892
    },
    {
      "epoch": 1.8791287270130335,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 7893
    },
    {
      "epoch": 1.8793667797417126,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 7894
    },
    {
      "epoch": 1.8796048324703922,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 7895
    },
    {
      "epoch": 1.8798428851990716,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 7896
    },
    {
      "epoch": 1.880080937927751,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 7897
    },
    {
      "epoch": 1.8803189906564304,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 7898
    },
    {
      "epoch": 1.8805570433851098,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 7899
    },
    {
      "epoch": 1.8807950961137894,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 7900
    },
    {
      "epoch": 1.8807950961137894,
      "eval_runtime": 0.542,
      "eval_samples_per_second": 1.845,
      "eval_steps_per_second": 1.845,
      "step": 7900
    },
    {
      "epoch": 1.8810331488424685,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 7901
    },
    {
      "epoch": 1.8812712015711481,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 7902
    },
    {
      "epoch": 1.8815092542998273,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 7903
    },
    {
      "epoch": 1.881747307028507,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 7904
    },
    {
      "epoch": 1.881985359757186,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 7905
    },
    {
      "epoch": 1.8822234124858657,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 7906
    },
    {
      "epoch": 1.882461465214545,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 7907
    },
    {
      "epoch": 1.8826995179432244,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 7908
    },
    {
      "epoch": 1.8829375706719038,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 7909
    },
    {
      "epoch": 1.8831756234005832,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 7910
    },
    {
      "epoch": 1.8834136761292626,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 7911
    },
    {
      "epoch": 1.883651728857942,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 7912
    },
    {
      "epoch": 1.8838897815866216,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 7913
    },
    {
      "epoch": 1.8841278343153007,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 7914
    },
    {
      "epoch": 1.8843658870439803,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 7915
    },
    {
      "epoch": 1.8846039397726595,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 7916
    },
    {
      "epoch": 1.884841992501339,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 7917
    },
    {
      "epoch": 1.8850800452300185,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 7918
    },
    {
      "epoch": 1.8853180979586979,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 7919
    },
    {
      "epoch": 1.8855561506873773,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 7920
    },
    {
      "epoch": 1.8857942034160566,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3875,
      "step": 7921
    },
    {
      "epoch": 1.886032256144736,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 7922
    },
    {
      "epoch": 1.8862703088734154,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 7923
    },
    {
      "epoch": 1.886508361602095,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3653,
      "step": 7924
    },
    {
      "epoch": 1.8867464143307742,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.313,
      "step": 7925
    },
    {
      "epoch": 1.8869844670594538,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 7926
    },
    {
      "epoch": 1.887222519788133,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 7927
    },
    {
      "epoch": 1.8874605725168125,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 7928
    },
    {
      "epoch": 1.887698625245492,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 7929
    },
    {
      "epoch": 1.8879366779741713,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 7930
    },
    {
      "epoch": 1.8881747307028507,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 7931
    },
    {
      "epoch": 1.88841278343153,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 7932
    },
    {
      "epoch": 1.8886508361602095,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 7933
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 7934
    },
    {
      "epoch": 1.8891269416175684,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 7935
    },
    {
      "epoch": 1.8893649943462476,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 7936
    },
    {
      "epoch": 1.8896030470749272,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2995,
      "step": 7937
    },
    {
      "epoch": 1.8898410998036064,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 7938
    },
    {
      "epoch": 1.890079152532286,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 7939
    },
    {
      "epoch": 1.8903172052609651,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 7940
    },
    {
      "epoch": 1.8905552579896447,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 7941
    },
    {
      "epoch": 1.8907933107183241,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 7942
    },
    {
      "epoch": 1.8910313634470035,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 7943
    },
    {
      "epoch": 1.891269416175683,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 7944
    },
    {
      "epoch": 1.8915074689043623,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3132,
      "step": 7945
    },
    {
      "epoch": 1.8917455216330419,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 7946
    },
    {
      "epoch": 1.891983574361721,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 7947
    },
    {
      "epoch": 1.8922216270904006,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 7948
    },
    {
      "epoch": 1.8924596798190798,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 7949
    },
    {
      "epoch": 1.8926977325477594,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 7950
    },
    {
      "epoch": 1.8929357852764386,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3706,
      "step": 7951
    },
    {
      "epoch": 1.8931738380051182,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 7952
    },
    {
      "epoch": 1.8934118907337976,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 7953
    },
    {
      "epoch": 1.893649943462477,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 7954
    },
    {
      "epoch": 1.8938879961911563,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.6092,
      "step": 7955
    },
    {
      "epoch": 1.8941260489198357,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2365,
      "step": 7956
    },
    {
      "epoch": 1.894364101648515,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 7957
    },
    {
      "epoch": 1.8946021543771945,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2823,
      "step": 7958
    },
    {
      "epoch": 1.894840207105874,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 7959
    },
    {
      "epoch": 1.8950782598345532,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 7960
    },
    {
      "epoch": 1.8953163125632329,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2842,
      "step": 7961
    },
    {
      "epoch": 1.895554365291912,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 7962
    },
    {
      "epoch": 1.8957924180205916,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.513,
      "step": 7963
    },
    {
      "epoch": 1.896030470749271,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 7964
    },
    {
      "epoch": 1.8962685234779504,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.6925,
      "step": 7965
    },
    {
      "epoch": 1.8965065762066298,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 7966
    },
    {
      "epoch": 1.8967446289353092,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 7967
    },
    {
      "epoch": 1.8969826816639885,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 7968
    },
    {
      "epoch": 1.897220734392668,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 7969
    },
    {
      "epoch": 1.8974587871213475,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 7970
    },
    {
      "epoch": 1.8976968398500267,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 7971
    },
    {
      "epoch": 1.8979348925787063,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 7972
    },
    {
      "epoch": 1.8981729453073855,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.6501,
      "step": 7973
    },
    {
      "epoch": 1.898410998036065,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 7974
    },
    {
      "epoch": 1.8986490507647444,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 7975
    },
    {
      "epoch": 1.8988871034934238,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3504,
      "step": 7976
    },
    {
      "epoch": 1.8991251562221032,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 7977
    },
    {
      "epoch": 1.8993632089507826,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 7978
    },
    {
      "epoch": 1.899601261679462,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 7979
    },
    {
      "epoch": 1.8998393144081414,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 7980
    },
    {
      "epoch": 1.900077367136821,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3197,
      "step": 7981
    },
    {
      "epoch": 1.9003154198655001,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.434,
      "step": 7982
    },
    {
      "epoch": 1.9005534725941797,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 7983
    },
    {
      "epoch": 1.9007915253228589,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 7984
    },
    {
      "epoch": 1.9010295780515385,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 7985
    },
    {
      "epoch": 1.9012676307802177,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3202,
      "step": 7986
    },
    {
      "epoch": 1.9015056835088973,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.299,
      "step": 7987
    },
    {
      "epoch": 1.9017437362375766,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 7988
    },
    {
      "epoch": 1.901981788966256,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 7989
    },
    {
      "epoch": 1.9022198416949354,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 7990
    },
    {
      "epoch": 1.9024578944236148,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 7991
    },
    {
      "epoch": 1.9026959471522944,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 7992
    },
    {
      "epoch": 1.9029339998809736,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 7993
    },
    {
      "epoch": 1.9031720526096532,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 7994
    },
    {
      "epoch": 1.9034101053383323,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 7995
    },
    {
      "epoch": 1.903648158067012,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 7996
    },
    {
      "epoch": 1.903886210795691,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.5432,
      "step": 7997
    },
    {
      "epoch": 1.9041242635243707,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 7998
    },
    {
      "epoch": 1.90436231625305,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3419,
      "step": 7999
    },
    {
      "epoch": 1.9046003689817295,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3757,
      "step": 8000
    },
    {
      "epoch": 1.9046003689817295,
      "eval_runtime": 0.5252,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 8000
    },
    {
      "epoch": 1.9048384217104088,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 8001
    },
    {
      "epoch": 1.9050764744390882,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 8002
    },
    {
      "epoch": 1.9053145271677676,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3297,
      "step": 8003
    },
    {
      "epoch": 1.905552579896447,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 8004
    },
    {
      "epoch": 1.9057906326251266,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 8005
    },
    {
      "epoch": 1.9060286853538058,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 8006
    },
    {
      "epoch": 1.9062667380824854,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3993,
      "step": 8007
    },
    {
      "epoch": 1.9065047908111645,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 8008
    },
    {
      "epoch": 1.9067428435398441,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 8009
    },
    {
      "epoch": 1.9069808962685235,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3019,
      "step": 8010
    },
    {
      "epoch": 1.907218948997203,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2882,
      "step": 8011
    },
    {
      "epoch": 1.9074570017258823,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 8012
    },
    {
      "epoch": 1.9076950544545617,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 8013
    },
    {
      "epoch": 1.907933107183241,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 8014
    },
    {
      "epoch": 1.9081711599119204,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 8015
    },
    {
      "epoch": 1.9084092126406,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 8016
    },
    {
      "epoch": 1.9086472653692792,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 8017
    },
    {
      "epoch": 1.9088853180979588,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 8018
    },
    {
      "epoch": 1.909123370826638,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 8019
    },
    {
      "epoch": 1.9093614235553176,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 8020
    },
    {
      "epoch": 1.909599476283997,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 8021
    },
    {
      "epoch": 1.9098375290126763,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 8022
    },
    {
      "epoch": 1.9100755817413557,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 8023
    },
    {
      "epoch": 1.910313634470035,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 8024
    },
    {
      "epoch": 1.9105516871987145,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 8025
    },
    {
      "epoch": 1.9107897399273939,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 8026
    },
    {
      "epoch": 1.9110277926560735,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 8027
    },
    {
      "epoch": 1.9112658453847526,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 8028
    },
    {
      "epoch": 1.9115038981134322,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 8029
    },
    {
      "epoch": 1.9117419508421114,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 8030
    },
    {
      "epoch": 1.911980003570791,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 8031
    },
    {
      "epoch": 1.9122180562994702,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4196,
      "step": 8032
    },
    {
      "epoch": 1.9124561090281498,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 8033
    },
    {
      "epoch": 1.9126941617568292,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 8034
    },
    {
      "epoch": 1.9129322144855085,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 8035
    },
    {
      "epoch": 1.913170267214188,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 8036
    },
    {
      "epoch": 1.9134083199428673,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 8037
    },
    {
      "epoch": 1.913646372671547,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 8038
    },
    {
      "epoch": 1.913884425400226,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4932,
      "step": 8039
    },
    {
      "epoch": 1.9141224781289057,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 8040
    },
    {
      "epoch": 1.9143605308575848,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 8041
    },
    {
      "epoch": 1.9145985835862644,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 8042
    },
    {
      "epoch": 1.9148366363149436,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 8043
    },
    {
      "epoch": 1.9150746890436232,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 8044
    },
    {
      "epoch": 1.9153127417723026,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.434,
      "step": 8045
    },
    {
      "epoch": 1.915550794500982,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 8046
    },
    {
      "epoch": 1.9157888472296614,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 8047
    },
    {
      "epoch": 1.9160268999583407,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3881,
      "step": 8048
    },
    {
      "epoch": 1.9162649526870201,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 8049
    },
    {
      "epoch": 1.9165030054156995,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 8050
    },
    {
      "epoch": 1.9167410581443791,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 8051
    },
    {
      "epoch": 1.9169791108730583,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 8052
    },
    {
      "epoch": 1.9172171636017379,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 8053
    },
    {
      "epoch": 1.917455216330417,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 8054
    },
    {
      "epoch": 1.9176932690590967,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 8055
    },
    {
      "epoch": 1.917931321787776,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 8056
    },
    {
      "epoch": 1.9181693745164554,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 8057
    },
    {
      "epoch": 1.9184074272451348,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 8058
    },
    {
      "epoch": 1.9186454799738142,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3911,
      "step": 8059
    },
    {
      "epoch": 1.9188835327024936,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3458,
      "step": 8060
    },
    {
      "epoch": 1.919121585431173,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 8061
    },
    {
      "epoch": 1.9193596381598526,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4127,
      "step": 8062
    },
    {
      "epoch": 1.9195976908885317,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 8063
    },
    {
      "epoch": 1.9198357436172113,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 8064
    },
    {
      "epoch": 1.9200737963458905,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 8065
    },
    {
      "epoch": 1.92031184907457,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 8066
    },
    {
      "epoch": 1.9205499018032495,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 8067
    },
    {
      "epoch": 1.9207879545319289,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 8068
    },
    {
      "epoch": 1.9210260072606082,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 8069
    },
    {
      "epoch": 1.9212640599892876,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 8070
    },
    {
      "epoch": 1.921502112717967,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 8071
    },
    {
      "epoch": 1.9217401654466464,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 8072
    },
    {
      "epoch": 1.921978218175326,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 8073
    },
    {
      "epoch": 1.9222162709040052,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 8074
    },
    {
      "epoch": 1.9224543236326848,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 8075
    },
    {
      "epoch": 1.922692376361364,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3965,
      "step": 8076
    },
    {
      "epoch": 1.9229304290900435,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 8077
    },
    {
      "epoch": 1.9231684818187227,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 8078
    },
    {
      "epoch": 1.9234065345474023,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 8079
    },
    {
      "epoch": 1.9236445872760817,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 8080
    },
    {
      "epoch": 1.923882640004761,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 8081
    },
    {
      "epoch": 1.9241206927334404,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2414,
      "step": 8082
    },
    {
      "epoch": 1.9243587454621198,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 8083
    },
    {
      "epoch": 1.9245967981907992,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3161,
      "step": 8084
    },
    {
      "epoch": 1.9248348509194786,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 8085
    },
    {
      "epoch": 1.9250729036481582,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 8086
    },
    {
      "epoch": 1.9253109563768374,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7671,
      "step": 8087
    },
    {
      "epoch": 1.925549009105517,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 8088
    },
    {
      "epoch": 1.9257870618341961,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 8089
    },
    {
      "epoch": 1.9260251145628757,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 8090
    },
    {
      "epoch": 1.9262631672915551,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 8091
    },
    {
      "epoch": 1.9265012200202345,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 8092
    },
    {
      "epoch": 1.9267392727489139,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 8093
    },
    {
      "epoch": 1.9269773254775933,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 8094
    },
    {
      "epoch": 1.9272153782062726,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2591,
      "step": 8095
    },
    {
      "epoch": 1.927453430934952,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 8096
    },
    {
      "epoch": 1.9276914836636316,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 8097
    },
    {
      "epoch": 1.9279295363923108,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2158,
      "step": 8098
    },
    {
      "epoch": 1.9281675891209904,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.7045,
      "step": 8099
    },
    {
      "epoch": 1.9284056418496696,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 8100
    },
    {
      "epoch": 1.9284056418496696,
      "eval_runtime": 0.5207,
      "eval_samples_per_second": 1.921,
      "eval_steps_per_second": 1.921,
      "step": 8100
    },
    {
      "epoch": 1.9286436945783492,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 8101
    },
    {
      "epoch": 1.9288817473070285,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2474,
      "step": 8102
    },
    {
      "epoch": 1.929119800035708,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 8103
    },
    {
      "epoch": 1.9293578527643873,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 8104
    },
    {
      "epoch": 1.9295959054930667,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 8105
    },
    {
      "epoch": 1.929833958221746,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2892,
      "step": 8106
    },
    {
      "epoch": 1.9300720109504255,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3928,
      "step": 8107
    },
    {
      "epoch": 1.930310063679105,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2673,
      "step": 8108
    },
    {
      "epoch": 1.9305481164077842,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3509,
      "step": 8109
    },
    {
      "epoch": 1.9307861691364638,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.3385,
      "step": 8110
    },
    {
      "epoch": 1.931024221865143,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 8111
    },
    {
      "epoch": 1.9312622745938226,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 8112
    },
    {
      "epoch": 1.931500327322502,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3061,
      "step": 8113
    },
    {
      "epoch": 1.9317383800511814,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 8114
    },
    {
      "epoch": 1.9319764327798608,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5125,
      "step": 8115
    },
    {
      "epoch": 1.9322144855085401,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 8116
    },
    {
      "epoch": 1.9324525382372195,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 8117
    },
    {
      "epoch": 1.932690590965899,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 8118
    },
    {
      "epoch": 1.9329286436945785,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 8119
    },
    {
      "epoch": 1.9331666964232577,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 8120
    },
    {
      "epoch": 1.9334047491519373,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 8121
    },
    {
      "epoch": 1.9336428018806164,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 8122
    },
    {
      "epoch": 1.933880854609296,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 8123
    },
    {
      "epoch": 1.9341189073379752,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 8124
    },
    {
      "epoch": 1.9343569600666548,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 8125
    },
    {
      "epoch": 1.9345950127953342,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 8126
    },
    {
      "epoch": 1.9348330655240136,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 8127
    },
    {
      "epoch": 1.935071118252693,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 8128
    },
    {
      "epoch": 1.9353091709813723,
      "grad_norm": 7.40625,
      "learning_rate": 3e-05,
      "loss": 2.4235,
      "step": 8129
    },
    {
      "epoch": 1.9355472237100517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 8130
    },
    {
      "epoch": 1.935785276438731,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3928,
      "step": 8131
    },
    {
      "epoch": 1.9360233291674107,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2728,
      "step": 8132
    },
    {
      "epoch": 1.9362613818960899,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 8133
    },
    {
      "epoch": 1.9364994346247695,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 8134
    },
    {
      "epoch": 1.9367374873534486,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 8135
    },
    {
      "epoch": 1.9369755400821282,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 8136
    },
    {
      "epoch": 1.9372135928108076,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 8137
    },
    {
      "epoch": 1.937451645539487,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 8138
    },
    {
      "epoch": 1.9376896982681664,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 8139
    },
    {
      "epoch": 1.9379277509968458,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.341,
      "step": 8140
    },
    {
      "epoch": 1.9381658037255252,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 8141
    },
    {
      "epoch": 1.9384038564542045,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 8142
    },
    {
      "epoch": 1.9386419091828841,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 8143
    },
    {
      "epoch": 1.9388799619115633,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 8144
    },
    {
      "epoch": 1.939118014640243,
      "grad_norm": 7.6875,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 8145
    },
    {
      "epoch": 1.939356067368922,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 8146
    },
    {
      "epoch": 1.9395941200976017,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4098,
      "step": 8147
    },
    {
      "epoch": 1.939832172826281,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.62,
      "step": 8148
    },
    {
      "epoch": 1.9400702255549604,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 2.6239,
      "step": 8149
    },
    {
      "epoch": 1.9403082782836398,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 8150
    },
    {
      "epoch": 1.9405463310123192,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 8151
    },
    {
      "epoch": 1.9407843837409986,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 8152
    },
    {
      "epoch": 1.941022436469678,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 8153
    },
    {
      "epoch": 1.9412604891983576,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.5488,
      "step": 8154
    },
    {
      "epoch": 1.9414985419270367,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 8155
    },
    {
      "epoch": 1.9417365946557164,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 8156
    },
    {
      "epoch": 1.9419746473843955,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 8157
    },
    {
      "epoch": 1.9422127001130751,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 8158
    },
    {
      "epoch": 1.9424507528417545,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 8159
    },
    {
      "epoch": 1.9426888055704339,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 8160
    },
    {
      "epoch": 1.9429268582991133,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 8161
    },
    {
      "epoch": 1.9431649110277927,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 8162
    },
    {
      "epoch": 1.943402963756472,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 8163
    },
    {
      "epoch": 1.9436410164851514,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 8164
    },
    {
      "epoch": 1.943879069213831,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3948,
      "step": 8165
    },
    {
      "epoch": 1.9441171219425102,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 8166
    },
    {
      "epoch": 1.9443551746711898,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.5334,
      "step": 8167
    },
    {
      "epoch": 1.944593227399869,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 8168
    },
    {
      "epoch": 1.9448312801285486,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.4268,
      "step": 8169
    },
    {
      "epoch": 1.9450693328572277,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3257,
      "step": 8170
    },
    {
      "epoch": 1.9453073855859073,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4035,
      "step": 8171
    },
    {
      "epoch": 1.9455454383145867,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 8172
    },
    {
      "epoch": 1.945783491043266,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 8173
    },
    {
      "epoch": 1.9460215437719455,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 8174
    },
    {
      "epoch": 1.9462595965006249,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 8175
    },
    {
      "epoch": 1.9464976492293042,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 8176
    },
    {
      "epoch": 1.9467357019579836,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 8177
    },
    {
      "epoch": 1.9469737546866632,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 8178
    },
    {
      "epoch": 1.9472118074153424,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4046,
      "step": 8179
    },
    {
      "epoch": 1.947449860144022,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 8180
    },
    {
      "epoch": 1.9476879128727012,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 8181
    },
    {
      "epoch": 1.9479259656013808,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 8182
    },
    {
      "epoch": 1.9481640183300601,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 8183
    },
    {
      "epoch": 1.9484020710587395,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2442,
      "step": 8184
    },
    {
      "epoch": 1.948640123787419,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 8185
    },
    {
      "epoch": 1.9488781765160983,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 8186
    },
    {
      "epoch": 1.9491162292447777,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 8187
    },
    {
      "epoch": 1.949354281973457,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 8188
    },
    {
      "epoch": 1.9495923347021367,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 8189
    },
    {
      "epoch": 1.9498303874308158,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 8190
    },
    {
      "epoch": 1.9500684401594954,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5069,
      "step": 8191
    },
    {
      "epoch": 1.9503064928881746,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 8192
    },
    {
      "epoch": 1.9505445456168542,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 8193
    },
    {
      "epoch": 1.9507825983455336,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 8194
    },
    {
      "epoch": 1.951020651074213,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 8195
    },
    {
      "epoch": 1.9512587038028923,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 8196
    },
    {
      "epoch": 1.9514967565315717,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 8197
    },
    {
      "epoch": 1.9517348092602511,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 8198
    },
    {
      "epoch": 1.9519728619889305,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 8199
    },
    {
      "epoch": 1.95221091471761,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 8200
    },
    {
      "epoch": 1.95221091471761,
      "eval_runtime": 0.5386,
      "eval_samples_per_second": 1.857,
      "eval_steps_per_second": 1.857,
      "step": 8200
    },
    {
      "epoch": 1.9524489674462893,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3253,
      "step": 8201
    },
    {
      "epoch": 1.9526870201749689,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 8202
    },
    {
      "epoch": 1.952925072903648,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3813,
      "step": 8203
    },
    {
      "epoch": 1.9531631256323276,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 8204
    },
    {
      "epoch": 1.953401178361007,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3381,
      "step": 8205
    },
    {
      "epoch": 1.9536392310896864,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 8206
    },
    {
      "epoch": 1.9538772838183658,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 8207
    },
    {
      "epoch": 1.9541153365470452,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 8208
    },
    {
      "epoch": 1.9543533892757246,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.5064,
      "step": 8209
    },
    {
      "epoch": 1.954591442004404,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5979,
      "step": 8210
    },
    {
      "epoch": 1.9548294947330835,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 8211
    },
    {
      "epoch": 1.9550675474617627,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 8212
    },
    {
      "epoch": 1.9553056001904423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.332,
      "step": 8213
    },
    {
      "epoch": 1.9555436529191215,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 8214
    },
    {
      "epoch": 1.955781705647801,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 8215
    },
    {
      "epoch": 1.9560197583764802,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 8216
    },
    {
      "epoch": 1.9562578111051598,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.4004,
      "step": 8217
    },
    {
      "epoch": 1.9564958638338392,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 8218
    },
    {
      "epoch": 1.9567339165625186,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5102,
      "step": 8219
    },
    {
      "epoch": 1.956971969291198,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 8220
    },
    {
      "epoch": 1.9572100220198774,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 8221
    },
    {
      "epoch": 1.9574480747485568,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5519,
      "step": 8222
    },
    {
      "epoch": 1.9576861274772361,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3923,
      "step": 8223
    },
    {
      "epoch": 1.9579241802059157,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3526,
      "step": 8224
    },
    {
      "epoch": 1.958162232934595,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 8225
    },
    {
      "epoch": 1.9584002856632745,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 8226
    },
    {
      "epoch": 1.9586383383919537,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 8227
    },
    {
      "epoch": 1.9588763911206333,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 8228
    },
    {
      "epoch": 1.9591144438493127,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4874,
      "step": 8229
    },
    {
      "epoch": 1.959352496577992,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 8230
    },
    {
      "epoch": 1.9595905493066714,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4206,
      "step": 8231
    },
    {
      "epoch": 1.9598286020353508,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 8232
    },
    {
      "epoch": 1.9600666547640302,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 8233
    },
    {
      "epoch": 1.9603047074927096,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 8234
    },
    {
      "epoch": 1.9605427602213892,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 8235
    },
    {
      "epoch": 1.9607808129500683,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 8236
    },
    {
      "epoch": 1.961018865678748,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 8237
    },
    {
      "epoch": 1.961256918407427,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 8238
    },
    {
      "epoch": 1.9614949711361067,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.2893,
      "step": 8239
    },
    {
      "epoch": 1.961733023864786,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 8240
    },
    {
      "epoch": 1.9619710765934655,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 8241
    },
    {
      "epoch": 1.9622091293221449,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 8242
    },
    {
      "epoch": 1.9624471820508242,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3393,
      "step": 8243
    },
    {
      "epoch": 1.9626852347795036,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2578,
      "step": 8244
    },
    {
      "epoch": 1.962923287508183,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 8245
    },
    {
      "epoch": 1.9631613402368626,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 8246
    },
    {
      "epoch": 1.9633993929655418,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 8247
    },
    {
      "epoch": 1.9636374456942214,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 8248
    },
    {
      "epoch": 1.9638754984229005,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 8249
    },
    {
      "epoch": 1.9641135511515802,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 8250
    },
    {
      "epoch": 1.9643516038802593,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 8251
    },
    {
      "epoch": 1.964589656608939,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 8252
    },
    {
      "epoch": 1.9648277093376183,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 8253
    },
    {
      "epoch": 1.9650657620662977,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 8254
    },
    {
      "epoch": 1.965303814794977,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 8255
    },
    {
      "epoch": 1.9655418675236564,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 8256
    },
    {
      "epoch": 1.965779920252336,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4961,
      "step": 8257
    },
    {
      "epoch": 1.9660179729810152,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 8258
    },
    {
      "epoch": 1.9662560257096948,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 8259
    },
    {
      "epoch": 1.966494078438374,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 8260
    },
    {
      "epoch": 1.9667321311670536,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 8261
    },
    {
      "epoch": 1.9669701838957327,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 8262
    },
    {
      "epoch": 1.9672082366244124,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 8263
    },
    {
      "epoch": 1.9674462893530917,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 8264
    },
    {
      "epoch": 1.9676843420817711,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 8265
    },
    {
      "epoch": 1.9679223948104505,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 8266
    },
    {
      "epoch": 1.9681604475391299,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 8267
    },
    {
      "epoch": 1.9683985002678093,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 8268
    },
    {
      "epoch": 1.9686365529964887,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5636,
      "step": 8269
    },
    {
      "epoch": 1.9688746057251683,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 8270
    },
    {
      "epoch": 1.9691126584538474,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 8271
    },
    {
      "epoch": 1.969350711182527,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 8272
    },
    {
      "epoch": 1.9695887639112062,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 8273
    },
    {
      "epoch": 1.9698268166398858,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 8274
    },
    {
      "epoch": 1.9700648693685652,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 8275
    },
    {
      "epoch": 1.9703029220972446,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 8276
    },
    {
      "epoch": 1.970540974825924,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 8277
    },
    {
      "epoch": 1.9707790275546033,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 8278
    },
    {
      "epoch": 1.9710170802832827,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.564,
      "step": 8279
    },
    {
      "epoch": 1.971255133011962,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2962,
      "step": 8280
    },
    {
      "epoch": 1.9714931857406417,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3927,
      "step": 8281
    },
    {
      "epoch": 1.9717312384693209,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 8282
    },
    {
      "epoch": 1.9719692911980005,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 8283
    },
    {
      "epoch": 1.9722073439266796,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 8284
    },
    {
      "epoch": 1.9724453966553592,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 8285
    },
    {
      "epoch": 1.9726834493840386,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 8286
    },
    {
      "epoch": 1.972921502112718,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 8287
    },
    {
      "epoch": 1.9731595548413974,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3543,
      "step": 8288
    },
    {
      "epoch": 1.9733976075700768,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 8289
    },
    {
      "epoch": 1.9736356602987561,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3481,
      "step": 8290
    },
    {
      "epoch": 1.9738737130274355,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 8291
    },
    {
      "epoch": 1.9741117657561151,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 8292
    },
    {
      "epoch": 1.9743498184847943,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 8293
    },
    {
      "epoch": 1.974587871213474,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 8294
    },
    {
      "epoch": 1.974825923942153,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 8295
    },
    {
      "epoch": 1.9750639766708327,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 8296
    },
    {
      "epoch": 1.9753020293995118,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 8297
    },
    {
      "epoch": 1.9755400821281914,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 8298
    },
    {
      "epoch": 1.9757781348568708,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3553,
      "step": 8299
    },
    {
      "epoch": 1.9760161875855502,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3713,
      "step": 8300
    },
    {
      "epoch": 1.9760161875855502,
      "eval_runtime": 0.5244,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 8300
    },
    {
      "epoch": 1.9762542403142296,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 8301
    },
    {
      "epoch": 1.976492293042909,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 8302
    },
    {
      "epoch": 1.9767303457715886,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3342,
      "step": 8303
    },
    {
      "epoch": 1.9769683985002677,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 8304
    },
    {
      "epoch": 1.9772064512289473,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 8305
    },
    {
      "epoch": 1.9774445039576265,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 8306
    },
    {
      "epoch": 1.977682556686306,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 8307
    },
    {
      "epoch": 1.9779206094149853,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 8308
    },
    {
      "epoch": 1.9781586621436649,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 8309
    },
    {
      "epoch": 1.9783967148723443,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.48,
      "step": 8310
    },
    {
      "epoch": 1.9786347676010236,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 8311
    },
    {
      "epoch": 1.978872820329703,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 8312
    },
    {
      "epoch": 1.9791108730583824,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4483,
      "step": 8313
    },
    {
      "epoch": 1.9793489257870618,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 8314
    },
    {
      "epoch": 1.9795869785157412,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 8315
    },
    {
      "epoch": 1.9798250312444208,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 8316
    },
    {
      "epoch": 1.9800630839731,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 8317
    },
    {
      "epoch": 1.9803011367017795,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 8318
    },
    {
      "epoch": 1.9805391894304587,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 8319
    },
    {
      "epoch": 1.9807772421591383,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 8320
    },
    {
      "epoch": 1.9810152948878177,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 8321
    },
    {
      "epoch": 1.981253347616497,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 8322
    },
    {
      "epoch": 1.9814914003451765,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 8323
    },
    {
      "epoch": 1.9817294530738558,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 8324
    },
    {
      "epoch": 1.9819675058025352,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 8325
    },
    {
      "epoch": 1.9822055585312146,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2882,
      "step": 8326
    },
    {
      "epoch": 1.9824436112598942,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1909,
      "step": 8327
    },
    {
      "epoch": 1.9826816639885734,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.5452,
      "step": 8328
    },
    {
      "epoch": 1.982919716717253,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 8329
    },
    {
      "epoch": 1.9831577694459321,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 8330
    },
    {
      "epoch": 1.9833958221746117,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2778,
      "step": 8331
    },
    {
      "epoch": 1.9836338749032911,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4067,
      "step": 8332
    },
    {
      "epoch": 1.9838719276319705,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 8333
    },
    {
      "epoch": 1.98410998036065,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4443,
      "step": 8334
    },
    {
      "epoch": 1.9843480330893293,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 8335
    },
    {
      "epoch": 1.9845860858180087,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4392,
      "step": 8336
    },
    {
      "epoch": 1.984824138546688,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 8337
    },
    {
      "epoch": 1.9850621912753676,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 8338
    },
    {
      "epoch": 1.9853002440040468,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5129,
      "step": 8339
    },
    {
      "epoch": 1.9855382967327264,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 8340
    },
    {
      "epoch": 1.9857763494614056,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 8341
    },
    {
      "epoch": 1.9860144021900852,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 8342
    },
    {
      "epoch": 1.9862524549187643,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 8343
    },
    {
      "epoch": 1.986490507647444,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 8344
    },
    {
      "epoch": 1.9867285603761233,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 8345
    },
    {
      "epoch": 1.9869666131048027,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 8346
    },
    {
      "epoch": 1.987204665833482,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 8347
    },
    {
      "epoch": 1.9874427185621615,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 8348
    },
    {
      "epoch": 1.987680771290841,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3151,
      "step": 8349
    },
    {
      "epoch": 1.9879188240195202,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 8350
    },
    {
      "epoch": 1.9881568767481999,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 8351
    },
    {
      "epoch": 1.988394929476879,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 8352
    },
    {
      "epoch": 1.9886329822055586,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 8353
    },
    {
      "epoch": 1.9888710349342378,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 8354
    },
    {
      "epoch": 1.9891090876629174,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 8355
    },
    {
      "epoch": 1.9893471403915968,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 8356
    },
    {
      "epoch": 1.9895851931202762,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3154,
      "step": 8357
    },
    {
      "epoch": 1.9898232458489555,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 8358
    },
    {
      "epoch": 1.990061298577635,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 8359
    },
    {
      "epoch": 1.9902993513063143,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 8360
    },
    {
      "epoch": 1.9905374040349937,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 8361
    },
    {
      "epoch": 1.9907754567636733,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 8362
    },
    {
      "epoch": 1.9910135094923525,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2387,
      "step": 8363
    },
    {
      "epoch": 1.991251562221032,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 8364
    },
    {
      "epoch": 1.9914896149497112,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 8365
    },
    {
      "epoch": 1.9917276676783908,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 8366
    },
    {
      "epoch": 1.9919657204070702,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 8367
    },
    {
      "epoch": 1.9922037731357496,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3076,
      "step": 8368
    },
    {
      "epoch": 1.992441825864429,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.377,
      "step": 8369
    },
    {
      "epoch": 1.9926798785931084,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 8370
    },
    {
      "epoch": 1.9929179313217877,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 8371
    },
    {
      "epoch": 1.9931559840504671,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 8372
    },
    {
      "epoch": 1.9933940367791467,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 8373
    },
    {
      "epoch": 1.9936320895078259,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 8374
    },
    {
      "epoch": 1.9938701422365055,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 8375
    },
    {
      "epoch": 1.9941081949651847,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3704,
      "step": 8376
    },
    {
      "epoch": 1.9943462476938643,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3446,
      "step": 8377
    },
    {
      "epoch": 1.9945843004225436,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3629,
      "step": 8378
    },
    {
      "epoch": 1.994822353151223,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 8379
    },
    {
      "epoch": 1.9950604058799024,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2542,
      "step": 8380
    },
    {
      "epoch": 1.9952984586085818,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 8381
    },
    {
      "epoch": 1.9955365113372612,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 8382
    },
    {
      "epoch": 1.9957745640659406,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3202,
      "step": 8383
    },
    {
      "epoch": 1.9960126167946202,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4572,
      "step": 8384
    },
    {
      "epoch": 1.9962506695232993,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 8385
    },
    {
      "epoch": 1.996488722251979,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 8386
    },
    {
      "epoch": 1.996726774980658,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 8387
    },
    {
      "epoch": 1.9969648277093377,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 8388
    },
    {
      "epoch": 1.9972028804380169,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 8389
    },
    {
      "epoch": 1.9974409331666965,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 8390
    },
    {
      "epoch": 1.9976789858953758,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 8391
    },
    {
      "epoch": 1.9979170386240552,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 8392
    },
    {
      "epoch": 1.9981550913527346,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4502,
      "step": 8393
    },
    {
      "epoch": 1.998393144081414,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3745,
      "step": 8394
    },
    {
      "epoch": 1.9986311968100936,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 8395
    },
    {
      "epoch": 1.9988692495387728,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 8396
    },
    {
      "epoch": 1.9991073022674524,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 8397
    },
    {
      "epoch": 1.9993453549961315,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 8398
    },
    {
      "epoch": 1.9995834077248111,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 8399
    },
    {
      "epoch": 1.9998214604534903,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 8400
    },
    {
      "epoch": 1.9998214604534903,
      "eval_runtime": 0.522,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 8400
    },
    {
      "epoch": 2.0002380527286796,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 4.5192,
      "step": 8401
    },
    {
      "epoch": 2.0004761054573588,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 8402
    },
    {
      "epoch": 2.0007141581860384,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 8403
    },
    {
      "epoch": 2.0009522109147175,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 8404
    },
    {
      "epoch": 2.001190263643397,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3575,
      "step": 8405
    },
    {
      "epoch": 2.0014283163720763,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 8406
    },
    {
      "epoch": 2.001666369100756,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4038,
      "step": 8407
    },
    {
      "epoch": 2.001904421829435,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 8408
    },
    {
      "epoch": 2.0021424745581147,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 8409
    },
    {
      "epoch": 2.002380527286794,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 8410
    },
    {
      "epoch": 2.0026185800154734,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 8411
    },
    {
      "epoch": 2.002856632744153,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 8412
    },
    {
      "epoch": 2.003094685472832,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 8413
    },
    {
      "epoch": 2.003332738201512,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 8414
    },
    {
      "epoch": 2.003570790930191,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 8415
    },
    {
      "epoch": 2.0038088436588706,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 8416
    },
    {
      "epoch": 2.0040468963875497,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 8417
    },
    {
      "epoch": 2.0042849491162293,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 8418
    },
    {
      "epoch": 2.0045230018449085,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 8419
    },
    {
      "epoch": 2.004761054573588,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 8420
    },
    {
      "epoch": 2.0049991073022673,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4316,
      "step": 8421
    },
    {
      "epoch": 2.005237160030947,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 8422
    },
    {
      "epoch": 2.005475212759626,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 8423
    },
    {
      "epoch": 2.0057132654883056,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 8424
    },
    {
      "epoch": 2.0059513182169852,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 8425
    },
    {
      "epoch": 2.0061893709456644,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 8426
    },
    {
      "epoch": 2.006427423674344,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3905,
      "step": 8427
    },
    {
      "epoch": 2.006665476403023,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 8428
    },
    {
      "epoch": 2.006903529131703,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 8429
    },
    {
      "epoch": 2.007141581860382,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 8430
    },
    {
      "epoch": 2.0073796345890615,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 8431
    },
    {
      "epoch": 2.0076176873177407,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 8432
    },
    {
      "epoch": 2.0078557400464203,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 8433
    },
    {
      "epoch": 2.0080937927750995,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 8434
    },
    {
      "epoch": 2.008331845503779,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 8435
    },
    {
      "epoch": 2.0085698982324587,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 8436
    },
    {
      "epoch": 2.008807950961138,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 8437
    },
    {
      "epoch": 2.0090460036898175,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 8438
    },
    {
      "epoch": 2.0092840564184966,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4118,
      "step": 8439
    },
    {
      "epoch": 2.009522109147176,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 8440
    },
    {
      "epoch": 2.0097601618758554,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 8441
    },
    {
      "epoch": 2.009998214604535,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 8442
    },
    {
      "epoch": 2.010236267333214,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 8443
    },
    {
      "epoch": 2.0104743200618937,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3639,
      "step": 8444
    },
    {
      "epoch": 2.010712372790573,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 8445
    },
    {
      "epoch": 2.0109504255192525,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 8446
    },
    {
      "epoch": 2.011188478247932,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 8447
    },
    {
      "epoch": 2.0114265309766113,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 8448
    },
    {
      "epoch": 2.011664583705291,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 8449
    },
    {
      "epoch": 2.01190263643397,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 8450
    },
    {
      "epoch": 2.0121406891626497,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.361,
      "step": 8451
    },
    {
      "epoch": 2.012378741891329,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 8452
    },
    {
      "epoch": 2.0126167946200084,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 8453
    },
    {
      "epoch": 2.0128548473486876,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 8454
    },
    {
      "epoch": 2.013092900077367,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 8455
    },
    {
      "epoch": 2.0133309528060463,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 8456
    },
    {
      "epoch": 2.013569005534726,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 8457
    },
    {
      "epoch": 2.0138070582634056,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 8458
    },
    {
      "epoch": 2.0140451109920847,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 8459
    },
    {
      "epoch": 2.0142831637207643,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 8460
    },
    {
      "epoch": 2.0145212164494435,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 8461
    },
    {
      "epoch": 2.014759269178123,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7456,
      "step": 8462
    },
    {
      "epoch": 2.0149973219068023,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 8463
    },
    {
      "epoch": 2.015235374635482,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 8464
    },
    {
      "epoch": 2.015473427364161,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2747,
      "step": 8465
    },
    {
      "epoch": 2.0157114800928406,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 8466
    },
    {
      "epoch": 2.01594953282152,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 8467
    },
    {
      "epoch": 2.0161875855501994,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 8468
    },
    {
      "epoch": 2.0164256382788786,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 8469
    },
    {
      "epoch": 2.016663691007558,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 8470
    },
    {
      "epoch": 2.0169017437362378,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 8471
    },
    {
      "epoch": 2.017139796464917,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 8472
    },
    {
      "epoch": 2.0173778491935965,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 8473
    },
    {
      "epoch": 2.0176159019222757,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 8474
    },
    {
      "epoch": 2.0178539546509553,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 8475
    },
    {
      "epoch": 2.0180920073796345,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 8476
    },
    {
      "epoch": 2.018330060108314,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 8477
    },
    {
      "epoch": 2.0185681128369932,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3421,
      "step": 8478
    },
    {
      "epoch": 2.018806165565673,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 8479
    },
    {
      "epoch": 2.019044218294352,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 8480
    },
    {
      "epoch": 2.0192822710230316,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 8481
    },
    {
      "epoch": 2.019520323751711,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 8482
    },
    {
      "epoch": 2.0197583764803904,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4436,
      "step": 8483
    },
    {
      "epoch": 2.01999642920907,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7036,
      "step": 8484
    },
    {
      "epoch": 2.020234481937749,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4265,
      "step": 8485
    },
    {
      "epoch": 2.0204725346664287,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 8486
    },
    {
      "epoch": 2.020710587395108,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 8487
    },
    {
      "epoch": 2.0209486401237875,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 8488
    },
    {
      "epoch": 2.0211866928524667,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 8489
    },
    {
      "epoch": 2.0214247455811463,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 8490
    },
    {
      "epoch": 2.0216627983098254,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 8491
    },
    {
      "epoch": 2.021900851038505,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 8492
    },
    {
      "epoch": 2.0221389037671846,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 8493
    },
    {
      "epoch": 2.022376956495864,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2485,
      "step": 8494
    },
    {
      "epoch": 2.0226150092245434,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 8495
    },
    {
      "epoch": 2.0228530619532226,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 8496
    },
    {
      "epoch": 2.023091114681902,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 8497
    },
    {
      "epoch": 2.0233291674105813,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.7741,
      "step": 8498
    },
    {
      "epoch": 2.023567220139261,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3386,
      "step": 8499
    },
    {
      "epoch": 2.02380527286794,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 8500
    },
    {
      "epoch": 2.02380527286794,
      "eval_runtime": 0.5245,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 8500
    },
    {
      "epoch": 2.0240433255966197,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 8501
    },
    {
      "epoch": 2.024281378325299,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 8502
    },
    {
      "epoch": 2.0245194310539785,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2428,
      "step": 8503
    },
    {
      "epoch": 2.024757483782658,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 8504
    },
    {
      "epoch": 2.0249955365113372,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3468,
      "step": 8505
    },
    {
      "epoch": 2.025233589240017,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 8506
    },
    {
      "epoch": 2.025471641968696,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3435,
      "step": 8507
    },
    {
      "epoch": 2.0257096946973756,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 8508
    },
    {
      "epoch": 2.0259477474260548,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 8509
    },
    {
      "epoch": 2.0261858001547344,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 8510
    },
    {
      "epoch": 2.0264238528834135,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 8511
    },
    {
      "epoch": 2.026661905612093,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2038,
      "step": 8512
    },
    {
      "epoch": 2.0268999583407723,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 8513
    },
    {
      "epoch": 2.027138011069452,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 8514
    },
    {
      "epoch": 2.027376063798131,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 8515
    },
    {
      "epoch": 2.0276141165268107,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 8516
    },
    {
      "epoch": 2.0278521692554903,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3721,
      "step": 8517
    },
    {
      "epoch": 2.0280902219841694,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 8518
    },
    {
      "epoch": 2.028328274712849,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 8519
    },
    {
      "epoch": 2.028566327441528,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 8520
    },
    {
      "epoch": 2.028804380170208,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 8521
    },
    {
      "epoch": 2.029042432898887,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 8522
    },
    {
      "epoch": 2.0292804856275666,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 8523
    },
    {
      "epoch": 2.0295185383562457,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 8524
    },
    {
      "epoch": 2.0297565910849253,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 8525
    },
    {
      "epoch": 2.0299946438136045,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 8526
    },
    {
      "epoch": 2.030232696542284,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3861,
      "step": 8527
    },
    {
      "epoch": 2.0304707492709637,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 8528
    },
    {
      "epoch": 2.030708801999643,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 8529
    },
    {
      "epoch": 2.0309468547283225,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 8530
    },
    {
      "epoch": 2.0311849074570016,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 8531
    },
    {
      "epoch": 2.0314229601856812,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 8532
    },
    {
      "epoch": 2.0316610129143604,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 8533
    },
    {
      "epoch": 2.03189906564304,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 8534
    },
    {
      "epoch": 2.032137118371719,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 8535
    },
    {
      "epoch": 2.032375171100399,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 8536
    },
    {
      "epoch": 2.032613223829078,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 8537
    },
    {
      "epoch": 2.0328512765577575,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3376,
      "step": 8538
    },
    {
      "epoch": 2.033089329286437,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3764,
      "step": 8539
    },
    {
      "epoch": 2.0333273820151163,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2244,
      "step": 8540
    },
    {
      "epoch": 2.033565434743796,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4251,
      "step": 8541
    },
    {
      "epoch": 2.033803487472475,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 8542
    },
    {
      "epoch": 2.0340415402011547,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3322,
      "step": 8543
    },
    {
      "epoch": 2.034279592929834,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3231,
      "step": 8544
    },
    {
      "epoch": 2.0345176456585135,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4389,
      "step": 8545
    },
    {
      "epoch": 2.0347556983871926,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 8546
    },
    {
      "epoch": 2.034993751115872,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 8547
    },
    {
      "epoch": 2.0352318038445514,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 8548
    },
    {
      "epoch": 2.035469856573231,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.34,
      "step": 8549
    },
    {
      "epoch": 2.0357079093019106,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 8550
    },
    {
      "epoch": 2.0359459620305898,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 8551
    },
    {
      "epoch": 2.0361840147592694,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3247,
      "step": 8552
    },
    {
      "epoch": 2.0364220674879485,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 8553
    },
    {
      "epoch": 2.036660120216628,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 8554
    },
    {
      "epoch": 2.0368981729453073,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3392,
      "step": 8555
    },
    {
      "epoch": 2.037136225673987,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4922,
      "step": 8556
    },
    {
      "epoch": 2.037374278402666,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.6859,
      "step": 8557
    },
    {
      "epoch": 2.0376123311313457,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 8558
    },
    {
      "epoch": 2.037850383860025,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 8559
    },
    {
      "epoch": 2.0380884365887044,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 8560
    },
    {
      "epoch": 2.0383264893173836,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 8561
    },
    {
      "epoch": 2.038564542046063,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 8562
    },
    {
      "epoch": 2.038802594774743,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 8563
    },
    {
      "epoch": 2.039040647503422,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 8564
    },
    {
      "epoch": 2.0392787002321016,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.6041,
      "step": 8565
    },
    {
      "epoch": 2.0395167529607807,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.546,
      "step": 8566
    },
    {
      "epoch": 2.0397548056894603,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 8567
    },
    {
      "epoch": 2.0399928584181395,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 8568
    },
    {
      "epoch": 2.040230911146819,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 8569
    },
    {
      "epoch": 2.0404689638754983,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.32,
      "step": 8570
    },
    {
      "epoch": 2.040707016604178,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 8571
    },
    {
      "epoch": 2.040945069332857,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 8572
    },
    {
      "epoch": 2.0411831220615366,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 8573
    },
    {
      "epoch": 2.0414211747902162,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2792,
      "step": 8574
    },
    {
      "epoch": 2.0416592275188954,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.3126,
      "step": 8575
    },
    {
      "epoch": 2.041897280247575,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 8576
    },
    {
      "epoch": 2.042135332976254,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 8577
    },
    {
      "epoch": 2.0423733857049338,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3545,
      "step": 8578
    },
    {
      "epoch": 2.042611438433613,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 8579
    },
    {
      "epoch": 2.0428494911622925,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 8580
    },
    {
      "epoch": 2.0430875438909717,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 8581
    },
    {
      "epoch": 2.0433255966196513,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 8582
    },
    {
      "epoch": 2.0435636493483305,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 8583
    },
    {
      "epoch": 2.04380170207701,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 8584
    },
    {
      "epoch": 2.0440397548056897,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 8585
    },
    {
      "epoch": 2.044277807534369,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 8586
    },
    {
      "epoch": 2.0445158602630484,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 8587
    },
    {
      "epoch": 2.0447539129917276,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 8588
    },
    {
      "epoch": 2.044991965720407,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 8589
    },
    {
      "epoch": 2.0452300184490864,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 8590
    },
    {
      "epoch": 2.045468071177766,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 8591
    },
    {
      "epoch": 2.045706123906445,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4463,
      "step": 8592
    },
    {
      "epoch": 2.0459441766351247,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 8593
    },
    {
      "epoch": 2.046182229363804,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 8594
    },
    {
      "epoch": 2.0464202820924835,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2823,
      "step": 8595
    },
    {
      "epoch": 2.046658334821163,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3542,
      "step": 8596
    },
    {
      "epoch": 2.0468963875498423,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 8597
    },
    {
      "epoch": 2.047134440278522,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 8598
    },
    {
      "epoch": 2.047372493007201,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 8599
    },
    {
      "epoch": 2.0476105457358806,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4157,
      "step": 8600
    },
    {
      "epoch": 2.0476105457358806,
      "eval_runtime": 0.5203,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 8600
    },
    {
      "epoch": 2.04784859846456,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 8601
    },
    {
      "epoch": 2.0480866511932394,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 8602
    },
    {
      "epoch": 2.0483247039219186,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.5373,
      "step": 8603
    },
    {
      "epoch": 2.048562756650598,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 8604
    },
    {
      "epoch": 2.0488008093792773,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 8605
    },
    {
      "epoch": 2.049038862107957,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3537,
      "step": 8606
    },
    {
      "epoch": 2.049276914836636,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 8607
    },
    {
      "epoch": 2.0495149675653157,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 8608
    },
    {
      "epoch": 2.0497530202939953,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 8609
    },
    {
      "epoch": 2.0499910730226745,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 8610
    },
    {
      "epoch": 2.050229125751354,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 8611
    },
    {
      "epoch": 2.0504671784800332,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 8612
    },
    {
      "epoch": 2.050705231208713,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 8613
    },
    {
      "epoch": 2.050943283937392,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 8614
    },
    {
      "epoch": 2.0511813366660716,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 8615
    },
    {
      "epoch": 2.0514193893947508,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 8616
    },
    {
      "epoch": 2.0516574421234304,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 8617
    },
    {
      "epoch": 2.0518954948521095,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3635,
      "step": 8618
    },
    {
      "epoch": 2.052133547580789,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 8619
    },
    {
      "epoch": 2.0523716003094687,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 8620
    },
    {
      "epoch": 2.052609653038148,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 8621
    },
    {
      "epoch": 2.0528477057668275,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 8622
    },
    {
      "epoch": 2.0530857584955067,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3697,
      "step": 8623
    },
    {
      "epoch": 2.0533238112241863,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 8624
    },
    {
      "epoch": 2.0535618639528654,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 8625
    },
    {
      "epoch": 2.053799916681545,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 8626
    },
    {
      "epoch": 2.054037969410224,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4382,
      "step": 8627
    },
    {
      "epoch": 2.054276022138904,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 8628
    },
    {
      "epoch": 2.054514074867583,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 8629
    },
    {
      "epoch": 2.0547521275962626,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 8630
    },
    {
      "epoch": 2.054990180324942,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 8631
    },
    {
      "epoch": 2.0552282330536213,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 8632
    },
    {
      "epoch": 2.055466285782301,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 8633
    },
    {
      "epoch": 2.05570433851098,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 8634
    },
    {
      "epoch": 2.0559423912396597,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 8635
    },
    {
      "epoch": 2.056180443968339,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 8636
    },
    {
      "epoch": 2.0564184966970185,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 8637
    },
    {
      "epoch": 2.0566565494256976,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 8638
    },
    {
      "epoch": 2.0568946021543772,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 8639
    },
    {
      "epoch": 2.0571326548830564,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 8640
    },
    {
      "epoch": 2.057370707611736,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 8641
    },
    {
      "epoch": 2.0576087603404156,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.283,
      "step": 8642
    },
    {
      "epoch": 2.057846813069095,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 8643
    },
    {
      "epoch": 2.0580848657977744,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 8644
    },
    {
      "epoch": 2.0583229185264535,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 8645
    },
    {
      "epoch": 2.058560971255133,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 8646
    },
    {
      "epoch": 2.0587990239838123,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 8647
    },
    {
      "epoch": 2.059037076712492,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 8648
    },
    {
      "epoch": 2.059275129441171,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 8649
    },
    {
      "epoch": 2.0595131821698507,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 8650
    },
    {
      "epoch": 2.05975123489853,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 8651
    },
    {
      "epoch": 2.0599892876272095,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 8652
    },
    {
      "epoch": 2.0602273403558886,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 8653
    },
    {
      "epoch": 2.060465393084568,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 8654
    },
    {
      "epoch": 2.060703445813248,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 8655
    },
    {
      "epoch": 2.060941498541927,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 8656
    },
    {
      "epoch": 2.0611795512706066,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 8657
    },
    {
      "epoch": 2.0614176039992858,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 8658
    },
    {
      "epoch": 2.0616556567279654,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 8659
    },
    {
      "epoch": 2.0618937094566445,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 8660
    },
    {
      "epoch": 2.062131762185324,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 8661
    },
    {
      "epoch": 2.0623698149140033,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.5856,
      "step": 8662
    },
    {
      "epoch": 2.062607867642683,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 8663
    },
    {
      "epoch": 2.062845920371362,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 8664
    },
    {
      "epoch": 2.0630839731000417,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.404,
      "step": 8665
    },
    {
      "epoch": 2.0633220258287213,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4249,
      "step": 8666
    },
    {
      "epoch": 2.0635600785574004,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 8667
    },
    {
      "epoch": 2.06379813128608,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 8668
    },
    {
      "epoch": 2.064036184014759,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.867,
      "step": 8669
    },
    {
      "epoch": 2.064274236743439,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4435,
      "step": 8670
    },
    {
      "epoch": 2.064512289472118,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 8671
    },
    {
      "epoch": 2.0647503422007976,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 8672
    },
    {
      "epoch": 2.0649883949294767,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 8673
    },
    {
      "epoch": 2.0652264476581563,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 8674
    },
    {
      "epoch": 2.0654645003868355,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 8675
    },
    {
      "epoch": 2.065702553115515,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 8676
    },
    {
      "epoch": 2.0659406058441947,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3174,
      "step": 8677
    },
    {
      "epoch": 2.066178658572874,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 8678
    },
    {
      "epoch": 2.0664167113015535,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 8679
    },
    {
      "epoch": 2.0666547640302326,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.7068,
      "step": 8680
    },
    {
      "epoch": 2.0668928167589122,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 8681
    },
    {
      "epoch": 2.0671308694875914,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 8682
    },
    {
      "epoch": 2.067368922216271,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 8683
    },
    {
      "epoch": 2.06760697494495,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 8684
    },
    {
      "epoch": 2.0678450276736298,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 8685
    },
    {
      "epoch": 2.068083080402309,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4083,
      "step": 8686
    },
    {
      "epoch": 2.0683211331309885,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 8687
    },
    {
      "epoch": 2.0685591858596677,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 8688
    },
    {
      "epoch": 2.0687972385883473,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 8689
    },
    {
      "epoch": 2.069035291317027,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 8690
    },
    {
      "epoch": 2.069273344045706,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 8691
    },
    {
      "epoch": 2.0695113967743857,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3295,
      "step": 8692
    },
    {
      "epoch": 2.069749449503065,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 8693
    },
    {
      "epoch": 2.0699875022317444,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 8694
    },
    {
      "epoch": 2.0702255549604236,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 8695
    },
    {
      "epoch": 2.070463607689103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8136,
      "step": 8696
    },
    {
      "epoch": 2.0707016604177824,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 8697
    },
    {
      "epoch": 2.070939713146462,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 8698
    },
    {
      "epoch": 2.071177765875141,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 8699
    },
    {
      "epoch": 2.0714158186038207,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3262,
      "step": 8700
    },
    {
      "epoch": 2.0714158186038207,
      "eval_runtime": 0.5216,
      "eval_samples_per_second": 1.917,
      "eval_steps_per_second": 1.917,
      "step": 8700
    },
    {
      "epoch": 2.0716538713325003,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 8701
    },
    {
      "epoch": 2.0718919240611795,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 8702
    },
    {
      "epoch": 2.072129976789859,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 8703
    },
    {
      "epoch": 2.0723680295185383,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3531,
      "step": 8704
    },
    {
      "epoch": 2.072606082247218,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 8705
    },
    {
      "epoch": 2.072844134975897,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 8706
    },
    {
      "epoch": 2.0730821877045766,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 8707
    },
    {
      "epoch": 2.073320240433256,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 8708
    },
    {
      "epoch": 2.0735582931619354,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 8709
    },
    {
      "epoch": 2.0737963458906146,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3817,
      "step": 8710
    },
    {
      "epoch": 2.074034398619294,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.438,
      "step": 8711
    },
    {
      "epoch": 2.074272451347974,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3125,
      "step": 8712
    },
    {
      "epoch": 2.074510504076653,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 8713
    },
    {
      "epoch": 2.0747485568053325,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 8714
    },
    {
      "epoch": 2.0749866095340117,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 8715
    },
    {
      "epoch": 2.0752246622626913,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 8716
    },
    {
      "epoch": 2.0754627149913705,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 8717
    },
    {
      "epoch": 2.07570076772005,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 8718
    },
    {
      "epoch": 2.0759388204487292,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 8719
    },
    {
      "epoch": 2.076176873177409,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 8720
    },
    {
      "epoch": 2.076414925906088,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 8721
    },
    {
      "epoch": 2.0766529786347676,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 8722
    },
    {
      "epoch": 2.076891031363447,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 8723
    },
    {
      "epoch": 2.0771290840921264,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 8724
    },
    {
      "epoch": 2.077367136820806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 8725
    },
    {
      "epoch": 2.077605189549485,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3023,
      "step": 8726
    },
    {
      "epoch": 2.0778432422781647,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 8727
    },
    {
      "epoch": 2.078081295006844,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3519,
      "step": 8728
    },
    {
      "epoch": 2.0783193477355235,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 8729
    },
    {
      "epoch": 2.0785574004642027,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 8730
    },
    {
      "epoch": 2.0787954531928823,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 8731
    },
    {
      "epoch": 2.0790335059215614,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 8732
    },
    {
      "epoch": 2.079271558650241,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 8733
    },
    {
      "epoch": 2.0795096113789207,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 8734
    },
    {
      "epoch": 2.0797476641076,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 8735
    },
    {
      "epoch": 2.0799857168362794,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 8736
    },
    {
      "epoch": 2.0802237695649586,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3836,
      "step": 8737
    },
    {
      "epoch": 2.080461822293638,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 8738
    },
    {
      "epoch": 2.0806998750223173,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3595,
      "step": 8739
    },
    {
      "epoch": 2.080937927750997,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4086,
      "step": 8740
    },
    {
      "epoch": 2.081175980479676,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 8741
    },
    {
      "epoch": 2.0814140332083557,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 8742
    },
    {
      "epoch": 2.081652085937035,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 8743
    },
    {
      "epoch": 2.0818901386657145,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3911,
      "step": 8744
    },
    {
      "epoch": 2.0821281913943936,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 8745
    },
    {
      "epoch": 2.0823662441230733,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 8746
    },
    {
      "epoch": 2.082604296851753,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 8747
    },
    {
      "epoch": 2.082842349580432,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 8748
    },
    {
      "epoch": 2.0830804023091116,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 8749
    },
    {
      "epoch": 2.083318455037791,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 8750
    },
    {
      "epoch": 2.0835565077664704,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 8751
    },
    {
      "epoch": 2.0837945604951496,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 8752
    },
    {
      "epoch": 2.084032613223829,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 8753
    },
    {
      "epoch": 2.0842706659525083,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 8754
    },
    {
      "epoch": 2.084508718681188,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 8755
    },
    {
      "epoch": 2.084746771409867,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 8756
    },
    {
      "epoch": 2.0849848241385467,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 8757
    },
    {
      "epoch": 2.0852228768672263,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 8758
    },
    {
      "epoch": 2.0854609295959055,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 8759
    },
    {
      "epoch": 2.085698982324585,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 8760
    },
    {
      "epoch": 2.085937035053264,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 8761
    },
    {
      "epoch": 2.086175087781944,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3164,
      "step": 8762
    },
    {
      "epoch": 2.086413140510623,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 8763
    },
    {
      "epoch": 2.0866511932393026,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 8764
    },
    {
      "epoch": 2.0868892459679818,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 8765
    },
    {
      "epoch": 2.0871272986966614,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 8766
    },
    {
      "epoch": 2.0873653514253405,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 8767
    },
    {
      "epoch": 2.08760340415402,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 8768
    },
    {
      "epoch": 2.0878414568826997,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 8769
    },
    {
      "epoch": 2.088079509611379,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3097,
      "step": 8770
    },
    {
      "epoch": 2.0883175623400585,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 8771
    },
    {
      "epoch": 2.0885556150687377,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 8772
    },
    {
      "epoch": 2.0887936677974173,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 8773
    },
    {
      "epoch": 2.0890317205260964,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3817,
      "step": 8774
    },
    {
      "epoch": 2.089269773254776,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 8775
    },
    {
      "epoch": 2.089507825983455,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 8776
    },
    {
      "epoch": 2.089745878712135,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 8777
    },
    {
      "epoch": 2.089983931440814,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 8778
    },
    {
      "epoch": 2.0902219841694936,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 8779
    },
    {
      "epoch": 2.0904600368981727,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 8780
    },
    {
      "epoch": 2.0906980896268523,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 8781
    },
    {
      "epoch": 2.090936142355532,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 8782
    },
    {
      "epoch": 2.091174195084211,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2547,
      "step": 8783
    },
    {
      "epoch": 2.0914122478128907,
      "grad_norm": 7.28125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 8784
    },
    {
      "epoch": 2.09165030054157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 8785
    },
    {
      "epoch": 2.0918883532702495,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 8786
    },
    {
      "epoch": 2.0921264059989286,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 8787
    },
    {
      "epoch": 2.0923644587276082,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 8788
    },
    {
      "epoch": 2.0926025114562874,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 8789
    },
    {
      "epoch": 2.092840564184967,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4487,
      "step": 8790
    },
    {
      "epoch": 2.093078616913646,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 8791
    },
    {
      "epoch": 2.0933166696423258,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5801,
      "step": 8792
    },
    {
      "epoch": 2.0935547223710054,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 8793
    },
    {
      "epoch": 2.0937927750996845,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 8794
    },
    {
      "epoch": 2.094030827828364,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8682,
      "step": 8795
    },
    {
      "epoch": 2.0942688805570433,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 8796
    },
    {
      "epoch": 2.094506933285723,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 8797
    },
    {
      "epoch": 2.094744986014402,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 8798
    },
    {
      "epoch": 2.0949830387430817,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 8799
    },
    {
      "epoch": 2.095221091471761,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4637,
      "step": 8800
    },
    {
      "epoch": 2.095221091471761,
      "eval_runtime": 0.5175,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 8800
    },
    {
      "epoch": 2.0954591442004404,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 8801
    },
    {
      "epoch": 2.0956971969291196,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 8802
    },
    {
      "epoch": 2.095935249657799,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3368,
      "step": 8803
    },
    {
      "epoch": 2.096173302386479,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.6666,
      "step": 8804
    },
    {
      "epoch": 2.096411355115158,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 8805
    },
    {
      "epoch": 2.0966494078438376,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 8806
    },
    {
      "epoch": 2.0968874605725167,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 8807
    },
    {
      "epoch": 2.0971255133011963,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 8808
    },
    {
      "epoch": 2.0973635660298755,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3666,
      "step": 8809
    },
    {
      "epoch": 2.097601618758555,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 8810
    },
    {
      "epoch": 2.0978396714872343,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4259,
      "step": 8811
    },
    {
      "epoch": 2.098077724215914,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 8812
    },
    {
      "epoch": 2.098315776944593,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 8813
    },
    {
      "epoch": 2.0985538296732726,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 8814
    },
    {
      "epoch": 2.0987918824019522,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4159,
      "step": 8815
    },
    {
      "epoch": 2.0990299351306314,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 8816
    },
    {
      "epoch": 2.099267987859311,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 8817
    },
    {
      "epoch": 2.09950604058799,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 8818
    },
    {
      "epoch": 2.09974409331667,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 8819
    },
    {
      "epoch": 2.099982146045349,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 8820
    },
    {
      "epoch": 2.1002201987740285,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4754,
      "step": 8821
    },
    {
      "epoch": 2.1004582515027077,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 8822
    },
    {
      "epoch": 2.1006963042313873,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2891,
      "step": 8823
    },
    {
      "epoch": 2.1009343569600665,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4234,
      "step": 8824
    },
    {
      "epoch": 2.101172409688746,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 8825
    },
    {
      "epoch": 2.1014104624174257,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 8826
    },
    {
      "epoch": 2.101648515146105,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 8827
    },
    {
      "epoch": 2.1018865678747845,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 8828
    },
    {
      "epoch": 2.1021246206034636,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 8829
    },
    {
      "epoch": 2.102362673332143,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 8830
    },
    {
      "epoch": 2.1026007260608224,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 8831
    },
    {
      "epoch": 2.102838778789502,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 8832
    },
    {
      "epoch": 2.103076831518181,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 8833
    },
    {
      "epoch": 2.1033148842468607,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 8834
    },
    {
      "epoch": 2.10355293697554,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 8835
    },
    {
      "epoch": 2.1037909897042195,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 8836
    },
    {
      "epoch": 2.1040290424328987,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 8837
    },
    {
      "epoch": 2.1042670951615783,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 8838
    },
    {
      "epoch": 2.104505147890258,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 8839
    },
    {
      "epoch": 2.104743200618937,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.3629,
      "step": 8840
    },
    {
      "epoch": 2.1049812533476167,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 8841
    },
    {
      "epoch": 2.105219306076296,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 8842
    },
    {
      "epoch": 2.1054573588049754,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 8843
    },
    {
      "epoch": 2.1056954115336546,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 8844
    },
    {
      "epoch": 2.105933464262334,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 8845
    },
    {
      "epoch": 2.1061715169910133,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3433,
      "step": 8846
    },
    {
      "epoch": 2.106409569719693,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 8847
    },
    {
      "epoch": 2.106647622448372,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 8848
    },
    {
      "epoch": 2.1068856751770517,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4182,
      "step": 8849
    },
    {
      "epoch": 2.1071237279057313,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 8850
    },
    {
      "epoch": 2.1073617806344105,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 8851
    },
    {
      "epoch": 2.10759983336309,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 8852
    },
    {
      "epoch": 2.1078378860917693,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 8853
    },
    {
      "epoch": 2.108075938820449,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2365,
      "step": 8854
    },
    {
      "epoch": 2.108313991549128,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 8855
    },
    {
      "epoch": 2.1085520442778076,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 8856
    },
    {
      "epoch": 2.108790097006487,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3408,
      "step": 8857
    },
    {
      "epoch": 2.1090281497351664,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 8858
    },
    {
      "epoch": 2.1092662024638456,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 8859
    },
    {
      "epoch": 2.109504255192525,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 8860
    },
    {
      "epoch": 2.1097423079212048,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 8861
    },
    {
      "epoch": 2.109980360649884,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 8862
    },
    {
      "epoch": 2.1102184133785635,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3939,
      "step": 8863
    },
    {
      "epoch": 2.1104564661072427,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3217,
      "step": 8864
    },
    {
      "epoch": 2.1106945188359223,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 8865
    },
    {
      "epoch": 2.1109325715646015,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 8866
    },
    {
      "epoch": 2.111170624293281,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 8867
    },
    {
      "epoch": 2.1114086770219602,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 8868
    },
    {
      "epoch": 2.11164672975064,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 8869
    },
    {
      "epoch": 2.111884782479319,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 8870
    },
    {
      "epoch": 2.1121228352079986,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 8871
    },
    {
      "epoch": 2.1123608879366778,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 8872
    },
    {
      "epoch": 2.1125989406653574,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 8873
    },
    {
      "epoch": 2.112836993394037,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 8874
    },
    {
      "epoch": 2.113075046122716,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.3705,
      "step": 8875
    },
    {
      "epoch": 2.1133130988513957,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3709,
      "step": 8876
    },
    {
      "epoch": 2.113551151580075,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 8877
    },
    {
      "epoch": 2.1137892043087545,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 8878
    },
    {
      "epoch": 2.1140272570374337,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 8879
    },
    {
      "epoch": 2.1142653097661133,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 8880
    },
    {
      "epoch": 2.1145033624947924,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 8881
    },
    {
      "epoch": 2.114741415223472,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 8882
    },
    {
      "epoch": 2.114979467952151,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 8883
    },
    {
      "epoch": 2.115217520680831,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 8884
    },
    {
      "epoch": 2.1154555734095104,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 8885
    },
    {
      "epoch": 2.1156936261381896,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 8886
    },
    {
      "epoch": 2.115931678866869,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 8887
    },
    {
      "epoch": 2.1161697315955483,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 8888
    },
    {
      "epoch": 2.116407784324228,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 8889
    },
    {
      "epoch": 2.116645837052907,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 8890
    },
    {
      "epoch": 2.1168838897815867,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 8891
    },
    {
      "epoch": 2.117121942510266,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 8892
    },
    {
      "epoch": 2.1173599952389455,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 8893
    },
    {
      "epoch": 2.1175980479676246,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2631,
      "step": 8894
    },
    {
      "epoch": 2.1178361006963042,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 8895
    },
    {
      "epoch": 2.118074153424984,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 8896
    },
    {
      "epoch": 2.118312206153663,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 8897
    },
    {
      "epoch": 2.1185502588823426,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 8898
    },
    {
      "epoch": 2.1187883116110218,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 8899
    },
    {
      "epoch": 2.1190263643397014,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 8900
    },
    {
      "epoch": 2.1190263643397014,
      "eval_runtime": 0.5234,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 8900
    },
    {
      "epoch": 2.1192644170683805,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 8901
    },
    {
      "epoch": 2.11950246979706,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 8902
    },
    {
      "epoch": 2.1197405225257393,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 8903
    },
    {
      "epoch": 2.119978575254419,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 8904
    },
    {
      "epoch": 2.120216627983098,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 8905
    },
    {
      "epoch": 2.1204546807117777,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 8906
    },
    {
      "epoch": 2.120692733440457,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 8907
    },
    {
      "epoch": 2.1209307861691364,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 8908
    },
    {
      "epoch": 2.121168838897816,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 8909
    },
    {
      "epoch": 2.121406891626495,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 8910
    },
    {
      "epoch": 2.121644944355175,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 8911
    },
    {
      "epoch": 2.121882997083854,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 8912
    },
    {
      "epoch": 2.1221210498125336,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 8913
    },
    {
      "epoch": 2.1223591025412127,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 8914
    },
    {
      "epoch": 2.1225971552698923,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 8915
    },
    {
      "epoch": 2.1228352079985715,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3439,
      "step": 8916
    },
    {
      "epoch": 2.123073260727251,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 8917
    },
    {
      "epoch": 2.1233113134559307,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.488,
      "step": 8918
    },
    {
      "epoch": 2.12354936618461,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 8919
    },
    {
      "epoch": 2.1237874189132895,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.487,
      "step": 8920
    },
    {
      "epoch": 2.1240254716419686,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 8921
    },
    {
      "epoch": 2.1242635243706482,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 8922
    },
    {
      "epoch": 2.1245015770993274,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 8923
    },
    {
      "epoch": 2.124739629828007,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 8924
    },
    {
      "epoch": 2.124977682556686,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 8925
    },
    {
      "epoch": 2.125215735285366,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 8926
    },
    {
      "epoch": 2.125453788014045,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2495,
      "step": 8927
    },
    {
      "epoch": 2.1256918407427245,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 8928
    },
    {
      "epoch": 2.1259298934714037,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 8929
    },
    {
      "epoch": 2.1261679462000833,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3004,
      "step": 8930
    },
    {
      "epoch": 2.126405998928763,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 8931
    },
    {
      "epoch": 2.126644051657442,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 8932
    },
    {
      "epoch": 2.1268821043861217,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 8933
    },
    {
      "epoch": 2.127120157114801,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 8934
    },
    {
      "epoch": 2.1273582098434805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3891,
      "step": 8935
    },
    {
      "epoch": 2.1275962625721596,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 8936
    },
    {
      "epoch": 2.127834315300839,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 8937
    },
    {
      "epoch": 2.1280723680295184,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.6429,
      "step": 8938
    },
    {
      "epoch": 2.128310420758198,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 8939
    },
    {
      "epoch": 2.128548473486877,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 8940
    },
    {
      "epoch": 2.1287865262155568,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 8941
    },
    {
      "epoch": 2.129024578944236,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 8942
    },
    {
      "epoch": 2.1292626316729155,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 8943
    },
    {
      "epoch": 2.129500684401595,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 8944
    },
    {
      "epoch": 2.1297387371302743,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 8945
    },
    {
      "epoch": 2.129976789858954,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 8946
    },
    {
      "epoch": 2.130214842587633,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 8947
    },
    {
      "epoch": 2.1304528953163127,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4167,
      "step": 8948
    },
    {
      "epoch": 2.130690948044992,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 8949
    },
    {
      "epoch": 2.1309290007736714,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 8950
    },
    {
      "epoch": 2.1311670535023506,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5205,
      "step": 8951
    },
    {
      "epoch": 2.13140510623103,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 8952
    },
    {
      "epoch": 2.13164315895971,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 8953
    },
    {
      "epoch": 2.131881211688389,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 8954
    },
    {
      "epoch": 2.1321192644170686,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 8955
    },
    {
      "epoch": 2.1323573171457477,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 8956
    },
    {
      "epoch": 2.1325953698744273,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4252,
      "step": 8957
    },
    {
      "epoch": 2.1328334226031065,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 8958
    },
    {
      "epoch": 2.133071475331786,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 8959
    },
    {
      "epoch": 2.1333095280604653,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 8960
    },
    {
      "epoch": 2.133547580789145,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 8961
    },
    {
      "epoch": 2.133785633517824,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4062,
      "step": 8962
    },
    {
      "epoch": 2.1340236862465036,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 8963
    },
    {
      "epoch": 2.134261738975183,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 8964
    },
    {
      "epoch": 2.1344997917038624,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 8965
    },
    {
      "epoch": 2.134737844432542,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4412,
      "step": 8966
    },
    {
      "epoch": 2.134975897161221,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 8967
    },
    {
      "epoch": 2.1352139498899008,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 8968
    },
    {
      "epoch": 2.13545200261858,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 8969
    },
    {
      "epoch": 2.1356900553472595,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 8970
    },
    {
      "epoch": 2.1359281080759387,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 8971
    },
    {
      "epoch": 2.1361661608046183,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 8972
    },
    {
      "epoch": 2.1364042135332975,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 8973
    },
    {
      "epoch": 2.136642266261977,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 8974
    },
    {
      "epoch": 2.1368803189906562,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 8975
    },
    {
      "epoch": 2.137118371719336,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 8976
    },
    {
      "epoch": 2.1373564244480154,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 8977
    },
    {
      "epoch": 2.1375944771766946,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 8978
    },
    {
      "epoch": 2.137832529905374,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 8979
    },
    {
      "epoch": 2.1380705826340534,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3786,
      "step": 8980
    },
    {
      "epoch": 2.138308635362733,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 8981
    },
    {
      "epoch": 2.138546688091412,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 8982
    },
    {
      "epoch": 2.1387847408200917,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.238,
      "step": 8983
    },
    {
      "epoch": 2.139022793548771,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 8984
    },
    {
      "epoch": 2.1392608462774505,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 8985
    },
    {
      "epoch": 2.1394988990061297,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4929,
      "step": 8986
    },
    {
      "epoch": 2.1397369517348093,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 8987
    },
    {
      "epoch": 2.139975004463489,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 8988
    },
    {
      "epoch": 2.140213057192168,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2937,
      "step": 8989
    },
    {
      "epoch": 2.1404511099208476,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 8990
    },
    {
      "epoch": 2.140689162649527,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 8991
    },
    {
      "epoch": 2.1409272153782064,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 8992
    },
    {
      "epoch": 2.1411652681068856,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 8993
    },
    {
      "epoch": 2.141403320835565,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2969,
      "step": 8994
    },
    {
      "epoch": 2.1416413735642443,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 8995
    },
    {
      "epoch": 2.141879426292924,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5892,
      "step": 8996
    },
    {
      "epoch": 2.142117479021603,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.4526,
      "step": 8997
    },
    {
      "epoch": 2.1423555317502827,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 8998
    },
    {
      "epoch": 2.142593584478962,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3602,
      "step": 8999
    },
    {
      "epoch": 2.1428316372076415,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 9000
    },
    {
      "epoch": 2.1428316372076415,
      "eval_runtime": 0.5289,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 9000
    },
    {
      "epoch": 2.143069689936321,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 9001
    },
    {
      "epoch": 2.1433077426650002,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3868,
      "step": 9002
    },
    {
      "epoch": 2.14354579539368,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 9003
    },
    {
      "epoch": 2.143783848122359,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 9004
    },
    {
      "epoch": 2.1440219008510386,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 9005
    },
    {
      "epoch": 2.1442599535797178,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 9006
    },
    {
      "epoch": 2.1444980063083974,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 9007
    },
    {
      "epoch": 2.1447360590370765,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 9008
    },
    {
      "epoch": 2.144974111765756,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.5318,
      "step": 9009
    },
    {
      "epoch": 2.1452121644944357,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 9010
    },
    {
      "epoch": 2.145450217223115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 9011
    },
    {
      "epoch": 2.1456882699517945,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 9012
    },
    {
      "epoch": 2.1459263226804737,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 9013
    },
    {
      "epoch": 2.1461643754091533,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 9014
    },
    {
      "epoch": 2.1464024281378324,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 9015
    },
    {
      "epoch": 2.146640480866512,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 9016
    },
    {
      "epoch": 2.146878533595191,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3076,
      "step": 9017
    },
    {
      "epoch": 2.147116586323871,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3744,
      "step": 9018
    },
    {
      "epoch": 2.14735463905255,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 9019
    },
    {
      "epoch": 2.1475926917812296,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 9020
    },
    {
      "epoch": 2.1478307445099087,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 9021
    },
    {
      "epoch": 2.1480687972385883,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 9022
    },
    {
      "epoch": 2.148306849967268,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 9023
    },
    {
      "epoch": 2.148544902695947,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 9024
    },
    {
      "epoch": 2.1487829554246267,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 9025
    },
    {
      "epoch": 2.149021008153306,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 9026
    },
    {
      "epoch": 2.1492590608819855,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.4226,
      "step": 9027
    },
    {
      "epoch": 2.1494971136106646,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 9028
    },
    {
      "epoch": 2.1497351663393443,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 9029
    },
    {
      "epoch": 2.1499732190680234,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 9030
    },
    {
      "epoch": 2.150211271796703,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 9031
    },
    {
      "epoch": 2.150449324525382,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 9032
    },
    {
      "epoch": 2.150687377254062,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 9033
    },
    {
      "epoch": 2.150925429982741,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 9034
    },
    {
      "epoch": 2.1511634827114205,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4261,
      "step": 9035
    },
    {
      "epoch": 2.1514015354401,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 9036
    },
    {
      "epoch": 2.1516395881687793,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5282,
      "step": 9037
    },
    {
      "epoch": 2.151877640897459,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 9038
    },
    {
      "epoch": 2.152115693626138,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 9039
    },
    {
      "epoch": 2.1523537463548177,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 9040
    },
    {
      "epoch": 2.152591799083497,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 9041
    },
    {
      "epoch": 2.1528298518121765,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 9042
    },
    {
      "epoch": 2.1530679045408556,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 9043
    },
    {
      "epoch": 2.153305957269535,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 9044
    },
    {
      "epoch": 2.153544009998215,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 9045
    },
    {
      "epoch": 2.153782062726894,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4697,
      "step": 9046
    },
    {
      "epoch": 2.1540201154555736,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 9047
    },
    {
      "epoch": 2.1542581681842528,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3484,
      "step": 9048
    },
    {
      "epoch": 2.1544962209129324,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.4156,
      "step": 9049
    },
    {
      "epoch": 2.1547342736416115,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 9050
    },
    {
      "epoch": 2.154972326370291,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 9051
    },
    {
      "epoch": 2.1552103790989703,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 9052
    },
    {
      "epoch": 2.15544843182765,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 9053
    },
    {
      "epoch": 2.155686484556329,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 9054
    },
    {
      "epoch": 2.1559245372850087,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 9055
    },
    {
      "epoch": 2.156162590013688,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 9056
    },
    {
      "epoch": 2.1564006427423674,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 9057
    },
    {
      "epoch": 2.156638695471047,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 9058
    },
    {
      "epoch": 2.156876748199726,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4073,
      "step": 9059
    },
    {
      "epoch": 2.157114800928406,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 9060
    },
    {
      "epoch": 2.157352853657085,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 9061
    },
    {
      "epoch": 2.1575909063857646,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 9062
    },
    {
      "epoch": 2.1578289591144437,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 9063
    },
    {
      "epoch": 2.1580670118431233,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.6137,
      "step": 9064
    },
    {
      "epoch": 2.1583050645718025,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 9065
    },
    {
      "epoch": 2.158543117300482,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 9066
    },
    {
      "epoch": 2.1587811700291613,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 9067
    },
    {
      "epoch": 2.159019222757841,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 9068
    },
    {
      "epoch": 2.1592572754865205,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 9069
    },
    {
      "epoch": 2.1594953282151996,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.332,
      "step": 9070
    },
    {
      "epoch": 2.1597333809438792,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1886,
      "step": 9071
    },
    {
      "epoch": 2.1599714336725584,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 9072
    },
    {
      "epoch": 2.160209486401238,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 9073
    },
    {
      "epoch": 2.160447539129917,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 9074
    },
    {
      "epoch": 2.1606855918585968,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 9075
    },
    {
      "epoch": 2.160923644587276,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 9076
    },
    {
      "epoch": 2.1611616973159555,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4467,
      "step": 9077
    },
    {
      "epoch": 2.1613997500446347,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 9078
    },
    {
      "epoch": 2.1616378027733143,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 9079
    },
    {
      "epoch": 2.161875855501994,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 9080
    },
    {
      "epoch": 2.162113908230673,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 9081
    },
    {
      "epoch": 2.1623519609593527,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 9082
    },
    {
      "epoch": 2.162590013688032,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 9083
    },
    {
      "epoch": 2.1628280664167114,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4112,
      "step": 9084
    },
    {
      "epoch": 2.1630661191453906,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 9085
    },
    {
      "epoch": 2.16330417187407,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4268,
      "step": 9086
    },
    {
      "epoch": 2.1635422246027494,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 9087
    },
    {
      "epoch": 2.163780277331429,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 9088
    },
    {
      "epoch": 2.164018330060108,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 9089
    },
    {
      "epoch": 2.1642563827887877,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 9090
    },
    {
      "epoch": 2.164494435517467,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 9091
    },
    {
      "epoch": 2.1647324882461465,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 9092
    },
    {
      "epoch": 2.164970540974826,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 9093
    },
    {
      "epoch": 2.1652085937035053,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 9094
    },
    {
      "epoch": 2.165446646432185,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4297,
      "step": 9095
    },
    {
      "epoch": 2.165684699160864,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 9096
    },
    {
      "epoch": 2.1659227518895436,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 9097
    },
    {
      "epoch": 2.166160804618223,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 9098
    },
    {
      "epoch": 2.1663988573469024,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 9099
    },
    {
      "epoch": 2.1666369100755816,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 9100
    },
    {
      "epoch": 2.1666369100755816,
      "eval_runtime": 0.5092,
      "eval_samples_per_second": 1.964,
      "eval_steps_per_second": 1.964,
      "step": 9100
    },
    {
      "epoch": 2.166874962804261,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 9101
    },
    {
      "epoch": 2.167113015532941,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 9102
    },
    {
      "epoch": 2.16735106826162,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 9103
    },
    {
      "epoch": 2.1675891209902995,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 9104
    },
    {
      "epoch": 2.1678271737189787,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 9105
    },
    {
      "epoch": 2.1680652264476583,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 9106
    },
    {
      "epoch": 2.1683032791763375,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 9107
    },
    {
      "epoch": 2.168541331905017,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 9108
    },
    {
      "epoch": 2.1687793846336962,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 9109
    },
    {
      "epoch": 2.169017437362376,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 9110
    },
    {
      "epoch": 2.169255490091055,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.361,
      "step": 9111
    },
    {
      "epoch": 2.1694935428197346,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 9112
    },
    {
      "epoch": 2.1697315955484138,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2244,
      "step": 9113
    },
    {
      "epoch": 2.1699696482770934,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 9114
    },
    {
      "epoch": 2.170207701005773,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 9115
    },
    {
      "epoch": 2.170445753734452,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 9116
    },
    {
      "epoch": 2.1706838064631317,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 9117
    },
    {
      "epoch": 2.170921859191811,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 9118
    },
    {
      "epoch": 2.1711599119204905,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.7061,
      "step": 9119
    },
    {
      "epoch": 2.1713979646491697,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 9120
    },
    {
      "epoch": 2.1716360173778493,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 9121
    },
    {
      "epoch": 2.1718740701065284,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 9122
    },
    {
      "epoch": 2.172112122835208,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 9123
    },
    {
      "epoch": 2.172350175563887,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 9124
    },
    {
      "epoch": 2.172588228292567,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 9125
    },
    {
      "epoch": 2.172826281021246,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 9126
    },
    {
      "epoch": 2.1730643337499256,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 9127
    },
    {
      "epoch": 2.173302386478605,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 9128
    },
    {
      "epoch": 2.1735404392072843,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 9129
    },
    {
      "epoch": 2.173778491935964,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 9130
    },
    {
      "epoch": 2.174016544664643,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 9131
    },
    {
      "epoch": 2.1742545973933227,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 9132
    },
    {
      "epoch": 2.174492650122002,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 9133
    },
    {
      "epoch": 2.1747307028506815,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 9134
    },
    {
      "epoch": 2.1749687555793606,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 9135
    },
    {
      "epoch": 2.1752068083080403,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 9136
    },
    {
      "epoch": 2.17544486103672,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 9137
    },
    {
      "epoch": 2.175682913765399,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 9138
    },
    {
      "epoch": 2.1759209664940786,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 9139
    },
    {
      "epoch": 2.176159019222758,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 9140
    },
    {
      "epoch": 2.1763970719514374,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 9141
    },
    {
      "epoch": 2.1766351246801166,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 9142
    },
    {
      "epoch": 2.176873177408796,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 9143
    },
    {
      "epoch": 2.1771112301374753,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 9144
    },
    {
      "epoch": 2.177349282866155,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 9145
    },
    {
      "epoch": 2.177587335594834,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 9146
    },
    {
      "epoch": 2.1778253883235137,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 9147
    },
    {
      "epoch": 2.178063441052193,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 9148
    },
    {
      "epoch": 2.1783014937808725,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 9149
    },
    {
      "epoch": 2.178539546509552,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 9150
    },
    {
      "epoch": 2.178777599238231,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 9151
    },
    {
      "epoch": 2.179015651966911,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 9152
    },
    {
      "epoch": 2.17925370469559,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 9153
    },
    {
      "epoch": 2.1794917574242696,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 9154
    },
    {
      "epoch": 2.1797298101529488,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 9155
    },
    {
      "epoch": 2.1799678628816284,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 9156
    },
    {
      "epoch": 2.1802059156103075,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 9157
    },
    {
      "epoch": 2.180443968338987,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2647,
      "step": 9158
    },
    {
      "epoch": 2.1806820210676663,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4072,
      "step": 9159
    },
    {
      "epoch": 2.180920073796346,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 9160
    },
    {
      "epoch": 2.1811581265250255,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 9161
    },
    {
      "epoch": 2.1813961792537047,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 9162
    },
    {
      "epoch": 2.1816342319823843,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 9163
    },
    {
      "epoch": 2.1818722847110634,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 9164
    },
    {
      "epoch": 2.182110337439743,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 9165
    },
    {
      "epoch": 2.182348390168422,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3504,
      "step": 9166
    },
    {
      "epoch": 2.182586442897102,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 9167
    },
    {
      "epoch": 2.182824495625781,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 9168
    },
    {
      "epoch": 2.1830625483544606,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 9169
    },
    {
      "epoch": 2.1833006010831397,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 9170
    },
    {
      "epoch": 2.1835386538118193,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 9171
    },
    {
      "epoch": 2.183776706540499,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 9172
    },
    {
      "epoch": 2.184014759269178,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4818,
      "step": 9173
    },
    {
      "epoch": 2.1842528119978577,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 9174
    },
    {
      "epoch": 2.184490864726537,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 9175
    },
    {
      "epoch": 2.1847289174552165,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3272,
      "step": 9176
    },
    {
      "epoch": 2.1849669701838956,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 9177
    },
    {
      "epoch": 2.1852050229125752,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 9178
    },
    {
      "epoch": 2.1854430756412544,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 9179
    },
    {
      "epoch": 2.185681128369934,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 9180
    },
    {
      "epoch": 2.185919181098613,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 9181
    },
    {
      "epoch": 2.1861572338272928,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 9182
    },
    {
      "epoch": 2.186395286555972,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3476,
      "step": 9183
    },
    {
      "epoch": 2.1866333392846515,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 9184
    },
    {
      "epoch": 2.186871392013331,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 9185
    },
    {
      "epoch": 2.1871094447420103,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 9186
    },
    {
      "epoch": 2.18734749747069,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 9187
    },
    {
      "epoch": 2.187585550199369,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 9188
    },
    {
      "epoch": 2.1878236029280487,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 9189
    },
    {
      "epoch": 2.188061655656728,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 9190
    },
    {
      "epoch": 2.1882997083854074,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 9191
    },
    {
      "epoch": 2.1885377611140866,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 9192
    },
    {
      "epoch": 2.188775813842766,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 9193
    },
    {
      "epoch": 2.189013866571446,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 9194
    },
    {
      "epoch": 2.189251919300125,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5274,
      "step": 9195
    },
    {
      "epoch": 2.1894899720288046,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 9196
    },
    {
      "epoch": 2.1897280247574837,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 9197
    },
    {
      "epoch": 2.1899660774861633,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 9198
    },
    {
      "epoch": 2.1902041302148425,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 9199
    },
    {
      "epoch": 2.190442182943522,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 9200
    },
    {
      "epoch": 2.190442182943522,
      "eval_runtime": 0.5235,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 9200
    },
    {
      "epoch": 2.1906802356722013,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 9201
    },
    {
      "epoch": 2.190918288400881,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 9202
    },
    {
      "epoch": 2.19115634112956,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 9203
    },
    {
      "epoch": 2.1913943938582396,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 9204
    },
    {
      "epoch": 2.191632446586919,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2547,
      "step": 9205
    },
    {
      "epoch": 2.1918704993155984,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 9206
    },
    {
      "epoch": 2.192108552044278,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 9207
    },
    {
      "epoch": 2.192346604772957,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 9208
    },
    {
      "epoch": 2.192584657501637,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.3746,
      "step": 9209
    },
    {
      "epoch": 2.192822710230316,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 9210
    },
    {
      "epoch": 2.1930607629589955,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 9211
    },
    {
      "epoch": 2.1932988156876747,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 9212
    },
    {
      "epoch": 2.1935368684163543,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5415,
      "step": 9213
    },
    {
      "epoch": 2.1937749211450335,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 9214
    },
    {
      "epoch": 2.194012973873713,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 9215
    },
    {
      "epoch": 2.1942510266023922,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 9216
    },
    {
      "epoch": 2.194489079331072,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 9217
    },
    {
      "epoch": 2.194727132059751,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 9218
    },
    {
      "epoch": 2.1949651847884306,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 9219
    },
    {
      "epoch": 2.19520323751711,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4527,
      "step": 9220
    },
    {
      "epoch": 2.1954412902457894,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 9221
    },
    {
      "epoch": 2.195679342974469,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 9222
    },
    {
      "epoch": 2.195917395703148,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 9223
    },
    {
      "epoch": 2.1961554484318278,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 9224
    },
    {
      "epoch": 2.196393501160507,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 9225
    },
    {
      "epoch": 2.1966315538891865,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 9226
    },
    {
      "epoch": 2.1968696066178657,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 9227
    },
    {
      "epoch": 2.1971076593465453,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.293,
      "step": 9228
    },
    {
      "epoch": 2.197345712075225,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 9229
    },
    {
      "epoch": 2.197583764803904,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 9230
    },
    {
      "epoch": 2.1978218175325837,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 9231
    },
    {
      "epoch": 2.198059870261263,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 9232
    },
    {
      "epoch": 2.1982979229899424,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 9233
    },
    {
      "epoch": 2.1985359757186216,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.4944,
      "step": 9234
    },
    {
      "epoch": 2.198774028447301,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7592,
      "step": 9235
    },
    {
      "epoch": 2.1990120811759803,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 9236
    },
    {
      "epoch": 2.19925013390466,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 9237
    },
    {
      "epoch": 2.199488186633339,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 9238
    },
    {
      "epoch": 2.1997262393620187,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 9239
    },
    {
      "epoch": 2.199964292090698,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 9240
    },
    {
      "epoch": 2.2002023448193775,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 9241
    },
    {
      "epoch": 2.200440397548057,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 9242
    },
    {
      "epoch": 2.2006784502767363,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 9243
    },
    {
      "epoch": 2.200916503005416,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.456,
      "step": 9244
    },
    {
      "epoch": 2.201154555734095,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 9245
    },
    {
      "epoch": 2.2013926084627746,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4,
      "step": 9246
    },
    {
      "epoch": 2.201630661191454,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3202,
      "step": 9247
    },
    {
      "epoch": 2.2018687139201334,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 9248
    },
    {
      "epoch": 2.2021067666488126,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 9249
    },
    {
      "epoch": 2.202344819377492,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 9250
    },
    {
      "epoch": 2.2025828721061713,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 9251
    },
    {
      "epoch": 2.202820924834851,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 9252
    },
    {
      "epoch": 2.2030589775635305,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4109,
      "step": 9253
    },
    {
      "epoch": 2.2032970302922097,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 9254
    },
    {
      "epoch": 2.2035350830208893,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 9255
    },
    {
      "epoch": 2.2037731357495685,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 9256
    },
    {
      "epoch": 2.204011188478248,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 9257
    },
    {
      "epoch": 2.2042492412069272,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 9258
    },
    {
      "epoch": 2.204487293935607,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4265,
      "step": 9259
    },
    {
      "epoch": 2.204725346664286,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 9260
    },
    {
      "epoch": 2.2049633993929656,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 9261
    },
    {
      "epoch": 2.2052014521216448,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 9262
    },
    {
      "epoch": 2.2054395048503244,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 9263
    },
    {
      "epoch": 2.205677557579004,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 9264
    },
    {
      "epoch": 2.205915610307683,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4061,
      "step": 9265
    },
    {
      "epoch": 2.2061536630363627,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 9266
    },
    {
      "epoch": 2.206391715765042,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 9267
    },
    {
      "epoch": 2.2066297684937215,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 9268
    },
    {
      "epoch": 2.2068678212224007,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 9269
    },
    {
      "epoch": 2.2071058739510803,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 9270
    },
    {
      "epoch": 2.2073439266797594,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 9271
    },
    {
      "epoch": 2.207581979408439,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 9272
    },
    {
      "epoch": 2.207820032137118,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 9273
    },
    {
      "epoch": 2.208058084865798,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 9274
    },
    {
      "epoch": 2.208296137594477,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 9275
    },
    {
      "epoch": 2.2085341903231566,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 9276
    },
    {
      "epoch": 2.208772243051836,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 9277
    },
    {
      "epoch": 2.2090102957805153,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 9278
    },
    {
      "epoch": 2.209248348509195,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 9279
    },
    {
      "epoch": 2.209486401237874,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7256,
      "step": 9280
    },
    {
      "epoch": 2.2097244539665537,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 9281
    },
    {
      "epoch": 2.209962506695233,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 9282
    },
    {
      "epoch": 2.2102005594239125,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4185,
      "step": 9283
    },
    {
      "epoch": 2.2104386121525916,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4571,
      "step": 9284
    },
    {
      "epoch": 2.2106766648812712,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 9285
    },
    {
      "epoch": 2.2109147176099504,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4647,
      "step": 9286
    },
    {
      "epoch": 2.21115277033863,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 9287
    },
    {
      "epoch": 2.2113908230673096,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 9288
    },
    {
      "epoch": 2.2116288757959888,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 9289
    },
    {
      "epoch": 2.2118669285246684,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 9290
    },
    {
      "epoch": 2.2121049812533475,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.3209,
      "step": 9291
    },
    {
      "epoch": 2.212343033982027,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 9292
    },
    {
      "epoch": 2.2125810867107063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 9293
    },
    {
      "epoch": 2.212819139439386,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 9294
    },
    {
      "epoch": 2.213057192168065,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 9295
    },
    {
      "epoch": 2.2132952448967447,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 9296
    },
    {
      "epoch": 2.213533297625424,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 9297
    },
    {
      "epoch": 2.2137713503541034,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 9298
    },
    {
      "epoch": 2.214009403082783,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 9299
    },
    {
      "epoch": 2.214247455811462,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 9300
    },
    {
      "epoch": 2.214247455811462,
      "eval_runtime": 0.5372,
      "eval_samples_per_second": 1.861,
      "eval_steps_per_second": 1.861,
      "step": 9300
    },
    {
      "epoch": 2.214485508540142,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 9301
    },
    {
      "epoch": 2.214723561268821,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.4058,
      "step": 9302
    },
    {
      "epoch": 2.2149616139975006,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 9303
    },
    {
      "epoch": 2.2151996667261797,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 9304
    },
    {
      "epoch": 2.2154377194548593,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.3435,
      "step": 9305
    },
    {
      "epoch": 2.2156757721835385,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 9306
    },
    {
      "epoch": 2.215913824912218,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 9307
    },
    {
      "epoch": 2.2161518776408973,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 9308
    },
    {
      "epoch": 2.216389930369577,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 9309
    },
    {
      "epoch": 2.216627983098256,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 9310
    },
    {
      "epoch": 2.2168660358269356,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2933,
      "step": 9311
    },
    {
      "epoch": 2.2171040885556152,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 9312
    },
    {
      "epoch": 2.2173421412842944,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 9313
    },
    {
      "epoch": 2.217580194012974,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 9314
    },
    {
      "epoch": 2.217818246741653,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 9315
    },
    {
      "epoch": 2.218056299470333,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 9316
    },
    {
      "epoch": 2.218294352199012,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 9317
    },
    {
      "epoch": 2.2185324049276915,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.5656,
      "step": 9318
    },
    {
      "epoch": 2.2187704576563707,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3623,
      "step": 9319
    },
    {
      "epoch": 2.2190085103850503,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3488,
      "step": 9320
    },
    {
      "epoch": 2.21924656311373,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 9321
    },
    {
      "epoch": 2.219484615842409,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 9322
    },
    {
      "epoch": 2.2197226685710887,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 9323
    },
    {
      "epoch": 2.219960721299768,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 9324
    },
    {
      "epoch": 2.2201987740284475,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 9325
    },
    {
      "epoch": 2.2204368267571266,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 9326
    },
    {
      "epoch": 2.220674879485806,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 9327
    },
    {
      "epoch": 2.2209129322144854,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 9328
    },
    {
      "epoch": 2.221150984943165,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 9329
    },
    {
      "epoch": 2.221389037671844,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 9330
    },
    {
      "epoch": 2.2216270904005238,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 9331
    },
    {
      "epoch": 2.221865143129203,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4182,
      "step": 9332
    },
    {
      "epoch": 2.2221031958578825,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2808,
      "step": 9333
    },
    {
      "epoch": 2.222341248586562,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 9334
    },
    {
      "epoch": 2.2225793013152413,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 9335
    },
    {
      "epoch": 2.222817354043921,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 9336
    },
    {
      "epoch": 2.2230554067726,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 9337
    },
    {
      "epoch": 2.2232934595012797,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 9338
    },
    {
      "epoch": 2.223531512229959,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 9339
    },
    {
      "epoch": 2.2237695649586384,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 9340
    },
    {
      "epoch": 2.2240076176873176,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 9341
    },
    {
      "epoch": 2.224245670415997,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 9342
    },
    {
      "epoch": 2.2244837231446764,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 9343
    },
    {
      "epoch": 2.224721775873356,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2312,
      "step": 9344
    },
    {
      "epoch": 2.224959828602035,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 9345
    },
    {
      "epoch": 2.2251978813307147,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 9346
    },
    {
      "epoch": 2.2254359340593943,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 9347
    },
    {
      "epoch": 2.2256739867880735,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 9348
    },
    {
      "epoch": 2.225912039516753,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3359,
      "step": 9349
    },
    {
      "epoch": 2.2261500922454323,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 9350
    },
    {
      "epoch": 2.226388144974112,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 9351
    },
    {
      "epoch": 2.226626197702791,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 9352
    },
    {
      "epoch": 2.2268642504314706,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3058,
      "step": 9353
    },
    {
      "epoch": 2.22710230316015,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 9354
    },
    {
      "epoch": 2.2273403558888294,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 9355
    },
    {
      "epoch": 2.227578408617509,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 9356
    },
    {
      "epoch": 2.227816461346188,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 9357
    },
    {
      "epoch": 2.2280545140748678,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 9358
    },
    {
      "epoch": 2.228292566803547,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 9359
    },
    {
      "epoch": 2.2285306195322265,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.517,
      "step": 9360
    },
    {
      "epoch": 2.2287686722609057,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 9361
    },
    {
      "epoch": 2.2290067249895853,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.5288,
      "step": 9362
    },
    {
      "epoch": 2.2292447777182645,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 9363
    },
    {
      "epoch": 2.229482830446944,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.5497,
      "step": 9364
    },
    {
      "epoch": 2.2297208831756232,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 9365
    },
    {
      "epoch": 2.229958935904303,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 9366
    },
    {
      "epoch": 2.230196988632982,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 9367
    },
    {
      "epoch": 2.2304350413616616,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 9368
    },
    {
      "epoch": 2.230673094090341,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 9369
    },
    {
      "epoch": 2.2309111468190204,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 9370
    },
    {
      "epoch": 2.2311491995477,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 9371
    },
    {
      "epoch": 2.231387252276379,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 9372
    },
    {
      "epoch": 2.2316253050050587,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3027,
      "step": 9373
    },
    {
      "epoch": 2.231863357733738,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 9374
    },
    {
      "epoch": 2.2321014104624175,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 9375
    },
    {
      "epoch": 2.2323394631910967,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 9376
    },
    {
      "epoch": 2.2325775159197763,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 9377
    },
    {
      "epoch": 2.2328155686484554,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 9378
    },
    {
      "epoch": 2.233053621377135,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 9379
    },
    {
      "epoch": 2.2332916741058146,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 9380
    },
    {
      "epoch": 2.233529726834494,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 9381
    },
    {
      "epoch": 2.2337677795631734,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4388,
      "step": 9382
    },
    {
      "epoch": 2.2340058322918526,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 9383
    },
    {
      "epoch": 2.234243885020532,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 9384
    },
    {
      "epoch": 2.2344819377492113,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 9385
    },
    {
      "epoch": 2.234719990477891,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 9386
    },
    {
      "epoch": 2.23495804320657,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 9387
    },
    {
      "epoch": 2.2351960959352497,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 9388
    },
    {
      "epoch": 2.235434148663929,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3527,
      "step": 9389
    },
    {
      "epoch": 2.2356722013926085,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 9390
    },
    {
      "epoch": 2.235910254121288,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 9391
    },
    {
      "epoch": 2.2361483068499672,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 9392
    },
    {
      "epoch": 2.236386359578647,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 9393
    },
    {
      "epoch": 2.236624412307326,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 9394
    },
    {
      "epoch": 2.2368624650360056,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 9395
    },
    {
      "epoch": 2.2371005177646848,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 9396
    },
    {
      "epoch": 2.2373385704933644,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 9397
    },
    {
      "epoch": 2.2375766232220435,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 9398
    },
    {
      "epoch": 2.237814675950723,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 9399
    },
    {
      "epoch": 2.2380527286794023,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.4799,
      "step": 9400
    },
    {
      "epoch": 2.2380527286794023,
      "eval_runtime": 0.5322,
      "eval_samples_per_second": 1.879,
      "eval_steps_per_second": 1.879,
      "step": 9400
    },
    {
      "epoch": 2.238290781408082,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 9401
    },
    {
      "epoch": 2.238528834136761,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4366,
      "step": 9402
    },
    {
      "epoch": 2.2387668868654407,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 9403
    },
    {
      "epoch": 2.2390049395941203,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 9404
    },
    {
      "epoch": 2.2392429923227994,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 9405
    },
    {
      "epoch": 2.239481045051479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 9406
    },
    {
      "epoch": 2.239719097780158,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 9407
    },
    {
      "epoch": 2.239957150508838,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 9408
    },
    {
      "epoch": 2.240195203237517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.7548,
      "step": 9409
    },
    {
      "epoch": 2.2404332559661966,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 9410
    },
    {
      "epoch": 2.2406713086948757,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 9411
    },
    {
      "epoch": 2.2409093614235553,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 9412
    },
    {
      "epoch": 2.241147414152235,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 9413
    },
    {
      "epoch": 2.241385466880914,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4923,
      "step": 9414
    },
    {
      "epoch": 2.2416235196095937,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 9415
    },
    {
      "epoch": 2.241861572338273,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 9416
    },
    {
      "epoch": 2.2420996250669525,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 9417
    },
    {
      "epoch": 2.2423376777956316,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 9418
    },
    {
      "epoch": 2.2425757305243113,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2916,
      "step": 9419
    },
    {
      "epoch": 2.2428137832529904,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 9420
    },
    {
      "epoch": 2.24305183598167,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3513,
      "step": 9421
    },
    {
      "epoch": 2.243289888710349,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 9422
    },
    {
      "epoch": 2.243527941439029,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 9423
    },
    {
      "epoch": 2.243765994167708,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 9424
    },
    {
      "epoch": 2.2440040468963875,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 9425
    },
    {
      "epoch": 2.244242099625067,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 9426
    },
    {
      "epoch": 2.2444801523537463,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 9427
    },
    {
      "epoch": 2.244718205082426,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 9428
    },
    {
      "epoch": 2.244956257811105,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 9429
    },
    {
      "epoch": 2.2451943105397847,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 9430
    },
    {
      "epoch": 2.245432363268464,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 9431
    },
    {
      "epoch": 2.2456704159971435,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 9432
    },
    {
      "epoch": 2.2459084687258226,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 9433
    },
    {
      "epoch": 2.246146521454502,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 9434
    },
    {
      "epoch": 2.2463845741831814,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 9435
    },
    {
      "epoch": 2.246622626911861,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 9436
    },
    {
      "epoch": 2.24686067964054,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 9437
    },
    {
      "epoch": 2.2470987323692198,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 9438
    },
    {
      "epoch": 2.2473367850978994,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 9439
    },
    {
      "epoch": 2.2475748378265785,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 9440
    },
    {
      "epoch": 2.247812890555258,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 9441
    },
    {
      "epoch": 2.2480509432839373,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3433,
      "step": 9442
    },
    {
      "epoch": 2.248288996012617,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.7133,
      "step": 9443
    },
    {
      "epoch": 2.248527048741296,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 9444
    },
    {
      "epoch": 2.2487651014699757,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 9445
    },
    {
      "epoch": 2.249003154198655,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 9446
    },
    {
      "epoch": 2.2492412069273344,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 9447
    },
    {
      "epoch": 2.249479259656014,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2841,
      "step": 9448
    },
    {
      "epoch": 2.249717312384693,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 9449
    },
    {
      "epoch": 2.249955365113373,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 9450
    },
    {
      "epoch": 2.250193417842052,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 9451
    },
    {
      "epoch": 2.2504314705707316,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 9452
    },
    {
      "epoch": 2.2506695232994107,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 9453
    },
    {
      "epoch": 2.2509075760280903,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 9454
    },
    {
      "epoch": 2.2511456287567695,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 9455
    },
    {
      "epoch": 2.251383681485449,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 9456
    },
    {
      "epoch": 2.2516217342141283,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 9457
    },
    {
      "epoch": 2.251859786942808,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 9458
    },
    {
      "epoch": 2.252097839671487,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 9459
    },
    {
      "epoch": 2.2523358924001666,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 9460
    },
    {
      "epoch": 2.2525739451288462,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 9461
    },
    {
      "epoch": 2.2528119978575254,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 9462
    },
    {
      "epoch": 2.253050050586205,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 9463
    },
    {
      "epoch": 2.253288103314884,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 9464
    },
    {
      "epoch": 2.2535261560435638,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 9465
    },
    {
      "epoch": 2.253764208772243,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 9466
    },
    {
      "epoch": 2.2540022615009225,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 9467
    },
    {
      "epoch": 2.2542403142296017,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 9468
    },
    {
      "epoch": 2.2544783669582813,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 9469
    },
    {
      "epoch": 2.254716419686961,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 9470
    },
    {
      "epoch": 2.25495447241564,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 9471
    },
    {
      "epoch": 2.2551925251443192,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 9472
    },
    {
      "epoch": 2.255430577872999,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 9473
    },
    {
      "epoch": 2.2556686306016784,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 9474
    },
    {
      "epoch": 2.2559066833303576,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3817,
      "step": 9475
    },
    {
      "epoch": 2.256144736059037,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 9476
    },
    {
      "epoch": 2.2563827887877164,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 9477
    },
    {
      "epoch": 2.256620841516396,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3334,
      "step": 9478
    },
    {
      "epoch": 2.256858894245075,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 9479
    },
    {
      "epoch": 2.2570969469737547,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3745,
      "step": 9480
    },
    {
      "epoch": 2.257334999702434,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 9481
    },
    {
      "epoch": 2.2575730524311135,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 9482
    },
    {
      "epoch": 2.257811105159793,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 9483
    },
    {
      "epoch": 2.2580491578884723,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 9484
    },
    {
      "epoch": 2.258287210617152,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 9485
    },
    {
      "epoch": 2.258525263345831,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 9486
    },
    {
      "epoch": 2.2587633160745106,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 9487
    },
    {
      "epoch": 2.25900136880319,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 9488
    },
    {
      "epoch": 2.2592394215318694,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 9489
    },
    {
      "epoch": 2.2594774742605486,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 9490
    },
    {
      "epoch": 2.259715526989228,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3918,
      "step": 9491
    },
    {
      "epoch": 2.2599535797179073,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4198,
      "step": 9492
    },
    {
      "epoch": 2.260191632446587,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 9493
    },
    {
      "epoch": 2.260429685175266,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 9494
    },
    {
      "epoch": 2.2606677379039457,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 9495
    },
    {
      "epoch": 2.2609057906326253,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2657,
      "step": 9496
    },
    {
      "epoch": 2.2611438433613045,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 9497
    },
    {
      "epoch": 2.261381896089984,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 9498
    },
    {
      "epoch": 2.2616199488186632,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 9499
    },
    {
      "epoch": 2.261858001547343,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 9500
    },
    {
      "epoch": 2.261858001547343,
      "eval_runtime": 0.5089,
      "eval_samples_per_second": 1.965,
      "eval_steps_per_second": 1.965,
      "step": 9500
    },
    {
      "epoch": 2.262096054276022,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 9501
    },
    {
      "epoch": 2.2623341070047016,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 9502
    },
    {
      "epoch": 2.2625721597333808,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 9503
    },
    {
      "epoch": 2.2628102124620604,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 9504
    },
    {
      "epoch": 2.26304826519074,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 9505
    },
    {
      "epoch": 2.263286317919419,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 9506
    },
    {
      "epoch": 2.2635243706480983,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2958,
      "step": 9507
    },
    {
      "epoch": 2.263762423376778,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 9508
    },
    {
      "epoch": 2.2640004761054575,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 9509
    },
    {
      "epoch": 2.2642385288341367,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 9510
    },
    {
      "epoch": 2.2644765815628163,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 9511
    },
    {
      "epoch": 2.2647146342914954,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 9512
    },
    {
      "epoch": 2.264952687020175,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 9513
    },
    {
      "epoch": 2.265190739748854,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3428,
      "step": 9514
    },
    {
      "epoch": 2.265428792477534,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 9515
    },
    {
      "epoch": 2.265666845206213,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 9516
    },
    {
      "epoch": 2.2659048979348926,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 9517
    },
    {
      "epoch": 2.266142950663572,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 9518
    },
    {
      "epoch": 2.2663810033922513,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 9519
    },
    {
      "epoch": 2.266619056120931,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 9520
    },
    {
      "epoch": 2.26685710884961,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3102,
      "step": 9521
    },
    {
      "epoch": 2.2670951615782897,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 9522
    },
    {
      "epoch": 2.267333214306969,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 9523
    },
    {
      "epoch": 2.2675712670356485,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 9524
    },
    {
      "epoch": 2.2678093197643276,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 9525
    },
    {
      "epoch": 2.2680473724930073,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 9526
    },
    {
      "epoch": 2.2682854252216864,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 9527
    },
    {
      "epoch": 2.268523477950366,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 9528
    },
    {
      "epoch": 2.268761530679045,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3869,
      "step": 9529
    },
    {
      "epoch": 2.268999583407725,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 9530
    },
    {
      "epoch": 2.2692376361364044,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 9531
    },
    {
      "epoch": 2.2694756888650836,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 9532
    },
    {
      "epoch": 2.269713741593763,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 9533
    },
    {
      "epoch": 2.2699517943224423,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 9534
    },
    {
      "epoch": 2.270189847051122,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 9535
    },
    {
      "epoch": 2.270427899779801,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 9536
    },
    {
      "epoch": 2.2706659525084807,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 9537
    },
    {
      "epoch": 2.27090400523716,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 9538
    },
    {
      "epoch": 2.2711420579658395,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 9539
    },
    {
      "epoch": 2.271380110694519,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 9540
    },
    {
      "epoch": 2.2716181634231982,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 9541
    },
    {
      "epoch": 2.271856216151878,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 9542
    },
    {
      "epoch": 2.272094268880557,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 9543
    },
    {
      "epoch": 2.2723323216092366,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 9544
    },
    {
      "epoch": 2.2725703743379158,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 9545
    },
    {
      "epoch": 2.2728084270665954,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 9546
    },
    {
      "epoch": 2.2730464797952745,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 9547
    },
    {
      "epoch": 2.273284532523954,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 9548
    },
    {
      "epoch": 2.2735225852526333,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 9549
    },
    {
      "epoch": 2.273760637981313,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 9550
    },
    {
      "epoch": 2.273998690709992,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 9551
    },
    {
      "epoch": 2.2742367434386717,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.496,
      "step": 9552
    },
    {
      "epoch": 2.2744747961673513,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 9553
    },
    {
      "epoch": 2.2747128488960304,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 9554
    },
    {
      "epoch": 2.27495090162471,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 9555
    },
    {
      "epoch": 2.275188954353389,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 9556
    },
    {
      "epoch": 2.275427007082069,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 9557
    },
    {
      "epoch": 2.275665059810748,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 9558
    },
    {
      "epoch": 2.2759031125394276,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.5296,
      "step": 9559
    },
    {
      "epoch": 2.2761411652681067,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3003,
      "step": 9560
    },
    {
      "epoch": 2.2763792179967863,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 9561
    },
    {
      "epoch": 2.276617270725466,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 9562
    },
    {
      "epoch": 2.276855323454145,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 9563
    },
    {
      "epoch": 2.2770933761828243,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 9564
    },
    {
      "epoch": 2.277331428911504,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 9565
    },
    {
      "epoch": 2.2775694816401835,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4225,
      "step": 9566
    },
    {
      "epoch": 2.2778075343688626,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 9567
    },
    {
      "epoch": 2.2780455870975422,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3014,
      "step": 9568
    },
    {
      "epoch": 2.2782836398262214,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 9569
    },
    {
      "epoch": 2.278521692554901,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 9570
    },
    {
      "epoch": 2.27875974528358,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 9571
    },
    {
      "epoch": 2.2789977980122598,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 9572
    },
    {
      "epoch": 2.279235850740939,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 9573
    },
    {
      "epoch": 2.2794739034696185,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4712,
      "step": 9574
    },
    {
      "epoch": 2.279711956198298,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 9575
    },
    {
      "epoch": 2.2799500089269773,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 9576
    },
    {
      "epoch": 2.280188061655657,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 9577
    },
    {
      "epoch": 2.280426114384336,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 9578
    },
    {
      "epoch": 2.2806641671130157,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 9579
    },
    {
      "epoch": 2.280902219841695,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.6635,
      "step": 9580
    },
    {
      "epoch": 2.2811402725703744,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3318,
      "step": 9581
    },
    {
      "epoch": 2.2813783252990536,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 9582
    },
    {
      "epoch": 2.281616378027733,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 9583
    },
    {
      "epoch": 2.2818544307564124,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 9584
    },
    {
      "epoch": 2.282092483485092,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3021,
      "step": 9585
    },
    {
      "epoch": 2.282330536213771,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 9586
    },
    {
      "epoch": 2.2825685889424507,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 9587
    },
    {
      "epoch": 2.2828066416711303,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3989,
      "step": 9588
    },
    {
      "epoch": 2.2830446943998095,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 9589
    },
    {
      "epoch": 2.283282747128489,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 9590
    },
    {
      "epoch": 2.2835207998571683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 9591
    },
    {
      "epoch": 2.283758852585848,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 9592
    },
    {
      "epoch": 2.283996905314527,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 9593
    },
    {
      "epoch": 2.2842349580432066,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 9594
    },
    {
      "epoch": 2.284473010771886,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 9595
    },
    {
      "epoch": 2.2847110635005654,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 9596
    },
    {
      "epoch": 2.284949116229245,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 9597
    },
    {
      "epoch": 2.285187168957924,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 9598
    },
    {
      "epoch": 2.2854252216866033,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 9599
    },
    {
      "epoch": 2.285663274415283,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 9600
    },
    {
      "epoch": 2.285663274415283,
      "eval_runtime": 0.5341,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 1.872,
      "step": 9600
    },
    {
      "epoch": 2.2859013271439625,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 9601
    },
    {
      "epoch": 2.2861393798726417,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 9602
    },
    {
      "epoch": 2.2863774326013213,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 9603
    },
    {
      "epoch": 2.2866154853300005,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 9604
    },
    {
      "epoch": 2.28685353805868,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 9605
    },
    {
      "epoch": 2.2870915907873592,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 9606
    },
    {
      "epoch": 2.287329643516039,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 9607
    },
    {
      "epoch": 2.287567696244718,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 9608
    },
    {
      "epoch": 2.2878057489733976,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 9609
    },
    {
      "epoch": 2.288043801702077,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 9610
    },
    {
      "epoch": 2.2882818544307564,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 9611
    },
    {
      "epoch": 2.288519907159436,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 9612
    },
    {
      "epoch": 2.288757959888115,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 9613
    },
    {
      "epoch": 2.2889960126167948,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 9614
    },
    {
      "epoch": 2.289234065345474,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 9615
    },
    {
      "epoch": 2.2894721180741535,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 9616
    },
    {
      "epoch": 2.2897101708028327,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 9617
    },
    {
      "epoch": 2.2899482235315123,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.4717,
      "step": 9618
    },
    {
      "epoch": 2.2901862762601914,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 9619
    },
    {
      "epoch": 2.290424328988871,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 9620
    },
    {
      "epoch": 2.29066238171755,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 9621
    },
    {
      "epoch": 2.29090043444623,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 9622
    },
    {
      "epoch": 2.2911384871749094,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 9623
    },
    {
      "epoch": 2.2913765399035886,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 9624
    },
    {
      "epoch": 2.291614592632268,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 9625
    },
    {
      "epoch": 2.2918526453609473,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 9626
    },
    {
      "epoch": 2.292090698089627,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 9627
    },
    {
      "epoch": 2.292328750818306,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 9628
    },
    {
      "epoch": 2.2925668035469857,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 9629
    },
    {
      "epoch": 2.292804856275665,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3323,
      "step": 9630
    },
    {
      "epoch": 2.2930429090043445,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3322,
      "step": 9631
    },
    {
      "epoch": 2.293280961733024,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 9632
    },
    {
      "epoch": 2.2935190144617033,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 9633
    },
    {
      "epoch": 2.293757067190383,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 9634
    },
    {
      "epoch": 2.293995119919062,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 9635
    },
    {
      "epoch": 2.2942331726477416,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 9636
    },
    {
      "epoch": 2.294471225376421,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.3333,
      "step": 9637
    },
    {
      "epoch": 2.2947092781051004,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 9638
    },
    {
      "epoch": 2.2949473308337796,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4039,
      "step": 9639
    },
    {
      "epoch": 2.295185383562459,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3842,
      "step": 9640
    },
    {
      "epoch": 2.2954234362911383,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 9641
    },
    {
      "epoch": 2.295661489019818,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 9642
    },
    {
      "epoch": 2.295899541748497,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 9643
    },
    {
      "epoch": 2.2961375944771767,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 9644
    },
    {
      "epoch": 2.2963756472058563,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3192,
      "step": 9645
    },
    {
      "epoch": 2.2966136999345355,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3468,
      "step": 9646
    },
    {
      "epoch": 2.296851752663215,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 9647
    },
    {
      "epoch": 2.2970898053918942,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 9648
    },
    {
      "epoch": 2.297327858120574,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4511,
      "step": 9649
    },
    {
      "epoch": 2.297565910849253,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 9650
    },
    {
      "epoch": 2.2978039635779326,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 9651
    },
    {
      "epoch": 2.2980420163066118,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.356,
      "step": 9652
    },
    {
      "epoch": 2.2982800690352914,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 9653
    },
    {
      "epoch": 2.298518121763971,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 9654
    },
    {
      "epoch": 2.29875617449265,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.5695,
      "step": 9655
    },
    {
      "epoch": 2.2989942272213293,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 9656
    },
    {
      "epoch": 2.299232279950009,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 9657
    },
    {
      "epoch": 2.2994703326786885,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 9658
    },
    {
      "epoch": 2.2997083854073677,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 9659
    },
    {
      "epoch": 2.2999464381360473,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 9660
    },
    {
      "epoch": 2.3001844908647264,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3675,
      "step": 9661
    },
    {
      "epoch": 2.300422543593406,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4337,
      "step": 9662
    },
    {
      "epoch": 2.300660596322085,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3436,
      "step": 9663
    },
    {
      "epoch": 2.300898649050765,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3555,
      "step": 9664
    },
    {
      "epoch": 2.301136701779444,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 9665
    },
    {
      "epoch": 2.3013747545081236,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 9666
    },
    {
      "epoch": 2.301612807236803,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 9667
    },
    {
      "epoch": 2.3018508599654823,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 9668
    },
    {
      "epoch": 2.302088912694162,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 9669
    },
    {
      "epoch": 2.302326965422841,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.445,
      "step": 9670
    },
    {
      "epoch": 2.3025650181515207,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 9671
    },
    {
      "epoch": 2.3028030708802,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 9672
    },
    {
      "epoch": 2.3030411236088795,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 9673
    },
    {
      "epoch": 2.3032791763375586,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 9674
    },
    {
      "epoch": 2.3035172290662382,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 9675
    },
    {
      "epoch": 2.3037552817949174,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 9676
    },
    {
      "epoch": 2.303993334523597,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 9677
    },
    {
      "epoch": 2.304231387252276,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 9678
    },
    {
      "epoch": 2.3044694399809558,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 9679
    },
    {
      "epoch": 2.3047074927096354,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 9680
    },
    {
      "epoch": 2.3049455454383145,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 9681
    },
    {
      "epoch": 2.305183598166994,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 9682
    },
    {
      "epoch": 2.3054216508956733,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 9683
    },
    {
      "epoch": 2.305659703624353,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 9684
    },
    {
      "epoch": 2.305897756353032,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 9685
    },
    {
      "epoch": 2.3061358090817117,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 9686
    },
    {
      "epoch": 2.306373861810391,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 9687
    },
    {
      "epoch": 2.3066119145390704,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 9688
    },
    {
      "epoch": 2.30684996726775,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 9689
    },
    {
      "epoch": 2.307088019996429,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 9690
    },
    {
      "epoch": 2.3073260727251084,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 9691
    },
    {
      "epoch": 2.307564125453788,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 9692
    },
    {
      "epoch": 2.3078021781824676,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 9693
    },
    {
      "epoch": 2.3080402309111467,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3929,
      "step": 9694
    },
    {
      "epoch": 2.3082782836398263,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.5452,
      "step": 9695
    },
    {
      "epoch": 2.3085163363685055,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 9696
    },
    {
      "epoch": 2.308754389097185,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 9697
    },
    {
      "epoch": 2.3089924418258643,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2768,
      "step": 9698
    },
    {
      "epoch": 2.309230494554544,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 9699
    },
    {
      "epoch": 2.309468547283223,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 9700
    },
    {
      "epoch": 2.309468547283223,
      "eval_runtime": 0.5094,
      "eval_samples_per_second": 1.963,
      "eval_steps_per_second": 1.963,
      "step": 9700
    },
    {
      "epoch": 2.3097066000119026,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 9701
    },
    {
      "epoch": 2.3099446527405822,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 9702
    },
    {
      "epoch": 2.3101827054692614,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.6248,
      "step": 9703
    },
    {
      "epoch": 2.310420758197941,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 9704
    },
    {
      "epoch": 2.31065881092662,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 9705
    },
    {
      "epoch": 2.3108968636553,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 9706
    },
    {
      "epoch": 2.311134916383979,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 9707
    },
    {
      "epoch": 2.3113729691126585,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 9708
    },
    {
      "epoch": 2.3116110218413377,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 9709
    },
    {
      "epoch": 2.3118490745700173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 9710
    },
    {
      "epoch": 2.3120871272986965,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 9711
    },
    {
      "epoch": 2.312325180027376,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 9712
    },
    {
      "epoch": 2.3125632327560552,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 9713
    },
    {
      "epoch": 2.312801285484735,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 9714
    },
    {
      "epoch": 2.3130393382134145,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 9715
    },
    {
      "epoch": 2.3132773909420936,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 9716
    },
    {
      "epoch": 2.313515443670773,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 9717
    },
    {
      "epoch": 2.3137534963994524,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 9718
    },
    {
      "epoch": 2.313991549128132,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 9719
    },
    {
      "epoch": 2.314229601856811,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 9720
    },
    {
      "epoch": 2.3144676545854908,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 9721
    },
    {
      "epoch": 2.31470570731417,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 9722
    },
    {
      "epoch": 2.3149437600428495,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3791,
      "step": 9723
    },
    {
      "epoch": 2.315181812771529,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.4461,
      "step": 9724
    },
    {
      "epoch": 2.3154198655002083,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 9725
    },
    {
      "epoch": 2.315657918228888,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 9726
    },
    {
      "epoch": 2.315895970957567,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 9727
    },
    {
      "epoch": 2.3161340236862467,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.5005,
      "step": 9728
    },
    {
      "epoch": 2.316372076414926,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 9729
    },
    {
      "epoch": 2.3166101291436054,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 9730
    },
    {
      "epoch": 2.3168481818722846,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3012,
      "step": 9731
    },
    {
      "epoch": 2.317086234600964,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 9732
    },
    {
      "epoch": 2.3173242873296434,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 9733
    },
    {
      "epoch": 2.317562340058323,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3191,
      "step": 9734
    },
    {
      "epoch": 2.317800392787002,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 9735
    },
    {
      "epoch": 2.3180384455156817,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 9736
    },
    {
      "epoch": 2.3182764982443613,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 9737
    },
    {
      "epoch": 2.3185145509730405,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 9738
    },
    {
      "epoch": 2.31875260370172,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 9739
    },
    {
      "epoch": 2.3189906564303993,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 9740
    },
    {
      "epoch": 2.319228709159079,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 9741
    },
    {
      "epoch": 2.319466761887758,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 9742
    },
    {
      "epoch": 2.3197048146164376,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 9743
    },
    {
      "epoch": 2.319942867345117,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 9744
    },
    {
      "epoch": 2.3201809200737964,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 9745
    },
    {
      "epoch": 2.3204189728024756,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 9746
    },
    {
      "epoch": 2.320657025531155,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 9747
    },
    {
      "epoch": 2.3208950782598343,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 9748
    },
    {
      "epoch": 2.321133130988514,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 9749
    },
    {
      "epoch": 2.3213711837171935,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 9750
    },
    {
      "epoch": 2.3216092364458727,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8172,
      "step": 9751
    },
    {
      "epoch": 2.3218472891745523,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 9752
    },
    {
      "epoch": 2.3220853419032315,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3167,
      "step": 9753
    },
    {
      "epoch": 2.322323394631911,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 9754
    },
    {
      "epoch": 2.3225614473605902,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 9755
    },
    {
      "epoch": 2.32279950008927,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3482,
      "step": 9756
    },
    {
      "epoch": 2.323037552817949,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 9757
    },
    {
      "epoch": 2.3232756055466286,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 9758
    },
    {
      "epoch": 2.323513658275308,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 9759
    },
    {
      "epoch": 2.3237517110039874,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 9760
    },
    {
      "epoch": 2.323989763732667,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 9761
    },
    {
      "epoch": 2.324227816461346,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 9762
    },
    {
      "epoch": 2.3244658691900257,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 9763
    },
    {
      "epoch": 2.324703921918705,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 9764
    },
    {
      "epoch": 2.3249419746473845,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 9765
    },
    {
      "epoch": 2.3251800273760637,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 9766
    },
    {
      "epoch": 2.3254180801047433,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 9767
    },
    {
      "epoch": 2.3256561328334224,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 9768
    },
    {
      "epoch": 2.325894185562102,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 9769
    },
    {
      "epoch": 2.326132238290781,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 9770
    },
    {
      "epoch": 2.326370291019461,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 9771
    },
    {
      "epoch": 2.3266083437481404,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.25,
      "step": 9772
    },
    {
      "epoch": 2.3268463964768196,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 9773
    },
    {
      "epoch": 2.327084449205499,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 9774
    },
    {
      "epoch": 2.3273225019341783,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 9775
    },
    {
      "epoch": 2.327560554662858,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 9776
    },
    {
      "epoch": 2.327798607391537,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5121,
      "step": 9777
    },
    {
      "epoch": 2.3280366601202167,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 9778
    },
    {
      "epoch": 2.328274712848896,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 9779
    },
    {
      "epoch": 2.3285127655775755,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 9780
    },
    {
      "epoch": 2.328750818306255,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 9781
    },
    {
      "epoch": 2.3289888710349342,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 9782
    },
    {
      "epoch": 2.3292269237636134,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.237,
      "step": 9783
    },
    {
      "epoch": 2.329464976492293,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 9784
    },
    {
      "epoch": 2.3297030292209726,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 9785
    },
    {
      "epoch": 2.3299410819496518,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3683,
      "step": 9786
    },
    {
      "epoch": 2.3301791346783314,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 9787
    },
    {
      "epoch": 2.3304171874070105,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 9788
    },
    {
      "epoch": 2.33065524013569,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 9789
    },
    {
      "epoch": 2.3308932928643693,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3451,
      "step": 9790
    },
    {
      "epoch": 2.331131345593049,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 9791
    },
    {
      "epoch": 2.331369398321728,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 9792
    },
    {
      "epoch": 2.3316074510504077,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 9793
    },
    {
      "epoch": 2.3318455037790873,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 9794
    },
    {
      "epoch": 2.3320835565077664,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 9795
    },
    {
      "epoch": 2.332321609236446,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 9796
    },
    {
      "epoch": 2.332559661965125,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3006,
      "step": 9797
    },
    {
      "epoch": 2.332797714693805,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 9798
    },
    {
      "epoch": 2.333035767422484,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 9799
    },
    {
      "epoch": 2.3332738201511636,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 9800
    },
    {
      "epoch": 2.3332738201511636,
      "eval_runtime": 0.5299,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 9800
    },
    {
      "epoch": 2.3335118728798427,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 9801
    },
    {
      "epoch": 2.3337499256085223,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 9802
    },
    {
      "epoch": 2.3339879783372015,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 9803
    },
    {
      "epoch": 2.334226031065881,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3893,
      "step": 9804
    },
    {
      "epoch": 2.3344640837945603,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 9805
    },
    {
      "epoch": 2.33470213652324,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 9806
    },
    {
      "epoch": 2.3349401892519195,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 9807
    },
    {
      "epoch": 2.3351782419805986,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 9808
    },
    {
      "epoch": 2.3354162947092783,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 9809
    },
    {
      "epoch": 2.3356543474379574,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 9810
    },
    {
      "epoch": 2.335892400166637,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 9811
    },
    {
      "epoch": 2.336130452895316,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 9812
    },
    {
      "epoch": 2.336368505623996,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 9813
    },
    {
      "epoch": 2.336606558352675,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 9814
    },
    {
      "epoch": 2.3368446110813546,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3524,
      "step": 9815
    },
    {
      "epoch": 2.337082663810034,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 9816
    },
    {
      "epoch": 2.3373207165387133,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 9817
    },
    {
      "epoch": 2.3375587692673925,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 9818
    },
    {
      "epoch": 2.337796821996072,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 9819
    },
    {
      "epoch": 2.3380348747247517,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 9820
    },
    {
      "epoch": 2.338272927453431,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 9821
    },
    {
      "epoch": 2.3385109801821105,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 9822
    },
    {
      "epoch": 2.3387490329107896,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 9823
    },
    {
      "epoch": 2.338987085639469,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 9824
    },
    {
      "epoch": 2.3392251383681484,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 9825
    },
    {
      "epoch": 2.339463191096828,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 9826
    },
    {
      "epoch": 2.339701243825507,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 9827
    },
    {
      "epoch": 2.3399392965541868,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 9828
    },
    {
      "epoch": 2.3401773492828664,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 3.0686,
      "step": 9829
    },
    {
      "epoch": 2.3404154020115455,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 9830
    },
    {
      "epoch": 2.340653454740225,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 9831
    },
    {
      "epoch": 2.3408915074689043,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 9832
    },
    {
      "epoch": 2.341129560197584,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 9833
    },
    {
      "epoch": 2.341367612926263,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 9834
    },
    {
      "epoch": 2.3416056656549427,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5025,
      "step": 9835
    },
    {
      "epoch": 2.341843718383622,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 9836
    },
    {
      "epoch": 2.3420817711123014,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4548,
      "step": 9837
    },
    {
      "epoch": 2.3423198238409806,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 9838
    },
    {
      "epoch": 2.34255787656966,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 9839
    },
    {
      "epoch": 2.3427959292983394,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 9840
    },
    {
      "epoch": 2.343033982027019,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4515,
      "step": 9841
    },
    {
      "epoch": 2.3432720347556986,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 9842
    },
    {
      "epoch": 2.3435100874843777,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.434,
      "step": 9843
    },
    {
      "epoch": 2.3437481402130573,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 9844
    },
    {
      "epoch": 2.3439861929417365,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 9845
    },
    {
      "epoch": 2.344224245670416,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2501,
      "step": 9846
    },
    {
      "epoch": 2.3444622983990953,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 9847
    },
    {
      "epoch": 2.344700351127775,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 9848
    },
    {
      "epoch": 2.344938403856454,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 9849
    },
    {
      "epoch": 2.3451764565851336,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3454,
      "step": 9850
    },
    {
      "epoch": 2.3454145093138132,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 9851
    },
    {
      "epoch": 2.3456525620424924,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 9852
    },
    {
      "epoch": 2.345890614771172,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 9853
    },
    {
      "epoch": 2.346128667499851,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 9854
    },
    {
      "epoch": 2.3463667202285308,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 9855
    },
    {
      "epoch": 2.34660477295721,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 9856
    },
    {
      "epoch": 2.3468428256858895,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 9857
    },
    {
      "epoch": 2.3470808784145687,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 9858
    },
    {
      "epoch": 2.3473189311432483,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 9859
    },
    {
      "epoch": 2.3475569838719275,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4429,
      "step": 9860
    },
    {
      "epoch": 2.347795036600607,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 9861
    },
    {
      "epoch": 2.3480330893292862,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 9862
    },
    {
      "epoch": 2.348271142057966,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 9863
    },
    {
      "epoch": 2.3485091947866454,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 9864
    },
    {
      "epoch": 2.3487472475153246,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 9865
    },
    {
      "epoch": 2.348985300244004,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2781,
      "step": 9866
    },
    {
      "epoch": 2.3492233529726834,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 9867
    },
    {
      "epoch": 2.349461405701363,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 9868
    },
    {
      "epoch": 2.349699458430042,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 9869
    },
    {
      "epoch": 2.3499375111587217,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 9870
    },
    {
      "epoch": 2.350175563887401,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 9871
    },
    {
      "epoch": 2.3504136166160805,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 9872
    },
    {
      "epoch": 2.35065166934476,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 9873
    },
    {
      "epoch": 2.3508897220734393,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 9874
    },
    {
      "epoch": 2.3511277748021184,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3661,
      "step": 9875
    },
    {
      "epoch": 2.351365827530798,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 9876
    },
    {
      "epoch": 2.3516038802594776,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 9877
    },
    {
      "epoch": 2.351841932988157,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 9878
    },
    {
      "epoch": 2.3520799857168364,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 9879
    },
    {
      "epoch": 2.3523180384455156,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 9880
    },
    {
      "epoch": 2.352556091174195,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 9881
    },
    {
      "epoch": 2.3527941439028743,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 9882
    },
    {
      "epoch": 2.353032196631554,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 9883
    },
    {
      "epoch": 2.353270249360233,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1886,
      "step": 9884
    },
    {
      "epoch": 2.3535083020889127,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 9885
    },
    {
      "epoch": 2.3537463548175923,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 9886
    },
    {
      "epoch": 2.3539844075462715,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 9887
    },
    {
      "epoch": 2.354222460274951,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4553,
      "step": 9888
    },
    {
      "epoch": 2.3544605130036302,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 9889
    },
    {
      "epoch": 2.35469856573231,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 9890
    },
    {
      "epoch": 2.354936618460989,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 9891
    },
    {
      "epoch": 2.3551746711896686,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 9892
    },
    {
      "epoch": 2.3554127239183478,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 9893
    },
    {
      "epoch": 2.3556507766470274,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 9894
    },
    {
      "epoch": 2.3558888293757065,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 9895
    },
    {
      "epoch": 2.356126882104386,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4718,
      "step": 9896
    },
    {
      "epoch": 2.3563649348330653,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 9897
    },
    {
      "epoch": 2.356602987561745,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 9898
    },
    {
      "epoch": 2.3568410402904245,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 9899
    },
    {
      "epoch": 2.3570790930191037,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 9900
    },
    {
      "epoch": 2.3570790930191037,
      "eval_runtime": 0.5264,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 9900
    },
    {
      "epoch": 2.3573171457477833,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 9901
    },
    {
      "epoch": 2.3575551984764624,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 9902
    },
    {
      "epoch": 2.357793251205142,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 9903
    },
    {
      "epoch": 2.358031303933821,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 9904
    },
    {
      "epoch": 2.358269356662501,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 9905
    },
    {
      "epoch": 2.35850740939118,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7528,
      "step": 9906
    },
    {
      "epoch": 2.3587454621198596,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 9907
    },
    {
      "epoch": 2.358983514848539,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 9908
    },
    {
      "epoch": 2.3592215675772183,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 9909
    },
    {
      "epoch": 2.3594596203058975,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 9910
    },
    {
      "epoch": 2.359697673034577,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 9911
    },
    {
      "epoch": 2.3599357257632567,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 9912
    },
    {
      "epoch": 2.360173778491936,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 9913
    },
    {
      "epoch": 2.3604118312206155,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3006,
      "step": 9914
    },
    {
      "epoch": 2.3606498839492946,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 9915
    },
    {
      "epoch": 2.3608879366779743,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 9916
    },
    {
      "epoch": 2.3611259894066534,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 9917
    },
    {
      "epoch": 2.361364042135333,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.189,
      "step": 9918
    },
    {
      "epoch": 2.361602094864012,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 9919
    },
    {
      "epoch": 2.361840147592692,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 9920
    },
    {
      "epoch": 2.3620782003213714,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 9921
    },
    {
      "epoch": 2.3623162530500506,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 9922
    },
    {
      "epoch": 2.36255430577873,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 9923
    },
    {
      "epoch": 2.3627923585074093,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 9924
    },
    {
      "epoch": 2.363030411236089,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 9925
    },
    {
      "epoch": 2.363268463964768,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 9926
    },
    {
      "epoch": 2.3635065166934477,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 9927
    },
    {
      "epoch": 2.363744569422127,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 9928
    },
    {
      "epoch": 2.3639826221508065,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 9929
    },
    {
      "epoch": 2.3642206748794856,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 9930
    },
    {
      "epoch": 2.3644587276081652,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 9931
    },
    {
      "epoch": 2.3646967803368444,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3725,
      "step": 9932
    },
    {
      "epoch": 2.364934833065524,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3125,
      "step": 9933
    },
    {
      "epoch": 2.3651728857942036,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 9934
    },
    {
      "epoch": 2.3654109385228828,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 9935
    },
    {
      "epoch": 2.3656489912515624,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 9936
    },
    {
      "epoch": 2.3658870439802415,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 9937
    },
    {
      "epoch": 2.366125096708921,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 9938
    },
    {
      "epoch": 2.3663631494376003,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 9939
    },
    {
      "epoch": 2.36660120216628,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4077,
      "step": 9940
    },
    {
      "epoch": 2.366839254894959,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 9941
    },
    {
      "epoch": 2.3670773076236387,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 9942
    },
    {
      "epoch": 2.3673153603523183,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 9943
    },
    {
      "epoch": 2.3675534130809974,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3061,
      "step": 9944
    },
    {
      "epoch": 2.367791465809677,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 9945
    },
    {
      "epoch": 2.368029518538356,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 9946
    },
    {
      "epoch": 2.368267571267036,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4373,
      "step": 9947
    },
    {
      "epoch": 2.368505623995715,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 9948
    },
    {
      "epoch": 2.3687436767243946,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 9949
    },
    {
      "epoch": 2.3689817294530737,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 9950
    },
    {
      "epoch": 2.3692197821817533,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3795,
      "step": 9951
    },
    {
      "epoch": 2.3694578349104325,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 9952
    },
    {
      "epoch": 2.369695887639112,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 9953
    },
    {
      "epoch": 2.3699339403677913,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 9954
    },
    {
      "epoch": 2.370171993096471,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3161,
      "step": 9955
    },
    {
      "epoch": 2.3704100458251505,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 9956
    },
    {
      "epoch": 2.3706480985538296,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 9957
    },
    {
      "epoch": 2.3708861512825092,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 9958
    },
    {
      "epoch": 2.3711242040111884,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 9959
    },
    {
      "epoch": 2.371362256739868,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 9960
    },
    {
      "epoch": 2.371600309468547,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 9961
    },
    {
      "epoch": 2.3718383621972268,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3443,
      "step": 9962
    },
    {
      "epoch": 2.372076414925906,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 9963
    },
    {
      "epoch": 2.3723144676545855,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 9964
    },
    {
      "epoch": 2.372552520383265,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 9965
    },
    {
      "epoch": 2.3727905731119443,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 9966
    },
    {
      "epoch": 2.3730286258406235,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8547,
      "step": 9967
    },
    {
      "epoch": 2.373266678569303,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 9968
    },
    {
      "epoch": 2.3735047312979827,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 9969
    },
    {
      "epoch": 2.373742784026662,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 9970
    },
    {
      "epoch": 2.3739808367553414,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 9971
    },
    {
      "epoch": 2.3742188894840206,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 9972
    },
    {
      "epoch": 2.3744569422127,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3876,
      "step": 9973
    },
    {
      "epoch": 2.3746949949413794,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 9974
    },
    {
      "epoch": 2.374933047670059,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 9975
    },
    {
      "epoch": 2.375171100398738,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 9976
    },
    {
      "epoch": 2.3754091531274177,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 9977
    },
    {
      "epoch": 2.3756472058560973,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 9978
    },
    {
      "epoch": 2.3758852585847765,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 9979
    },
    {
      "epoch": 2.376123311313456,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 9980
    },
    {
      "epoch": 2.3763613640421353,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 9981
    },
    {
      "epoch": 2.376599416770815,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 9982
    },
    {
      "epoch": 2.376837469499494,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3833,
      "step": 9983
    },
    {
      "epoch": 2.3770755222281736,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 9984
    },
    {
      "epoch": 2.377313574956853,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 9985
    },
    {
      "epoch": 2.3775516276855324,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 9986
    },
    {
      "epoch": 2.3777896804142116,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2788,
      "step": 9987
    },
    {
      "epoch": 2.378027733142891,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 9988
    },
    {
      "epoch": 2.3782657858715703,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 9989
    },
    {
      "epoch": 2.37850383860025,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 9990
    },
    {
      "epoch": 2.3787418913289295,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 9991
    },
    {
      "epoch": 2.3789799440576087,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 9992
    },
    {
      "epoch": 2.3792179967862883,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 9993
    },
    {
      "epoch": 2.3794560495149675,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 9994
    },
    {
      "epoch": 2.379694102243647,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 9995
    },
    {
      "epoch": 2.3799321549723262,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 9996
    },
    {
      "epoch": 2.380170207701006,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 9997
    },
    {
      "epoch": 2.380408260429685,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 9998
    },
    {
      "epoch": 2.3806463131583646,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 9999
    },
    {
      "epoch": 2.380884365887044,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 10000
    },
    {
      "epoch": 2.380884365887044,
      "eval_runtime": 0.5218,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 10000
    },
    {
      "epoch": 2.3811224186157234,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 10001
    },
    {
      "epoch": 2.3813604713444025,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3941,
      "step": 10002
    },
    {
      "epoch": 2.381598524073082,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 10003
    },
    {
      "epoch": 2.3818365768017618,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 10004
    },
    {
      "epoch": 2.382074629530441,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.5786,
      "step": 10005
    },
    {
      "epoch": 2.3823126822591205,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 10006
    },
    {
      "epoch": 2.3825507349877997,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 10007
    },
    {
      "epoch": 2.3827887877164793,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 10008
    },
    {
      "epoch": 2.3830268404451584,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 10009
    },
    {
      "epoch": 2.383264893173838,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 10010
    },
    {
      "epoch": 2.383502945902517,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 10011
    },
    {
      "epoch": 2.383740998631197,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 10012
    },
    {
      "epoch": 2.3839790513598764,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 10013
    },
    {
      "epoch": 2.3842171040885556,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 10014
    },
    {
      "epoch": 2.384455156817235,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 10015
    },
    {
      "epoch": 2.3846932095459143,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 10016
    },
    {
      "epoch": 2.384931262274594,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 10017
    },
    {
      "epoch": 2.385169315003273,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 10018
    },
    {
      "epoch": 2.3854073677319527,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 10019
    },
    {
      "epoch": 2.385645420460632,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 10020
    },
    {
      "epoch": 2.3858834731893115,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 10021
    },
    {
      "epoch": 2.3861215259179906,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 10022
    },
    {
      "epoch": 2.3863595786466703,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4141,
      "step": 10023
    },
    {
      "epoch": 2.3865976313753494,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 10024
    },
    {
      "epoch": 2.386835684104029,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 10025
    },
    {
      "epoch": 2.3870737368327086,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4502,
      "step": 10026
    },
    {
      "epoch": 2.387311789561388,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6626,
      "step": 10027
    },
    {
      "epoch": 2.3875498422900674,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 10028
    },
    {
      "epoch": 2.3877878950187466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 10029
    },
    {
      "epoch": 2.388025947747426,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 10030
    },
    {
      "epoch": 2.3882640004761053,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 10031
    },
    {
      "epoch": 2.388502053204785,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 10032
    },
    {
      "epoch": 2.388740105933464,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 10033
    },
    {
      "epoch": 2.3889781586621437,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3841,
      "step": 10034
    },
    {
      "epoch": 2.3892162113908233,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4904,
      "step": 10035
    },
    {
      "epoch": 2.3894542641195025,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 10036
    },
    {
      "epoch": 2.389692316848182,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 10037
    },
    {
      "epoch": 2.3899303695768612,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 10038
    },
    {
      "epoch": 2.390168422305541,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 10039
    },
    {
      "epoch": 2.39040647503422,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 10040
    },
    {
      "epoch": 2.3906445277628996,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 10041
    },
    {
      "epoch": 2.3908825804915788,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 10042
    },
    {
      "epoch": 2.3911206332202584,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 10043
    },
    {
      "epoch": 2.3913586859489375,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 10044
    },
    {
      "epoch": 2.391596738677617,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 10045
    },
    {
      "epoch": 2.3918347914062963,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 10046
    },
    {
      "epoch": 2.392072844134976,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 10047
    },
    {
      "epoch": 2.3923108968636555,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 10048
    },
    {
      "epoch": 2.3925489495923347,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 10049
    },
    {
      "epoch": 2.3927870023210143,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 10050
    },
    {
      "epoch": 2.3930250550496934,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 10051
    },
    {
      "epoch": 2.393263107778373,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 10052
    },
    {
      "epoch": 2.393501160507052,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 10053
    },
    {
      "epoch": 2.393739213235732,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 10054
    },
    {
      "epoch": 2.393977265964411,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 10055
    },
    {
      "epoch": 2.3942153186930906,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 10056
    },
    {
      "epoch": 2.39445337142177,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 10057
    },
    {
      "epoch": 2.3946914241504493,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 10058
    },
    {
      "epoch": 2.3949294768791285,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 10059
    },
    {
      "epoch": 2.395167529607808,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 10060
    },
    {
      "epoch": 2.3954055823364877,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 10061
    },
    {
      "epoch": 2.395643635065167,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 10062
    },
    {
      "epoch": 2.3958816877938465,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 10063
    },
    {
      "epoch": 2.3961197405225256,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 10064
    },
    {
      "epoch": 2.3963577932512052,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 10065
    },
    {
      "epoch": 2.3965958459798844,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 10066
    },
    {
      "epoch": 2.396833898708564,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 10067
    },
    {
      "epoch": 2.397071951437243,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3588,
      "step": 10068
    },
    {
      "epoch": 2.3973100041659228,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.5364,
      "step": 10069
    },
    {
      "epoch": 2.3975480568946024,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 10070
    },
    {
      "epoch": 2.3977861096232815,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 10071
    },
    {
      "epoch": 2.398024162351961,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 10072
    },
    {
      "epoch": 2.3982622150806403,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 10073
    },
    {
      "epoch": 2.39850026780932,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 10074
    },
    {
      "epoch": 2.398738320537999,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7781,
      "step": 10075
    },
    {
      "epoch": 2.3989763732666787,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 10076
    },
    {
      "epoch": 2.399214425995358,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 10077
    },
    {
      "epoch": 2.3994524787240374,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 10078
    },
    {
      "epoch": 2.3996905314527166,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 10079
    },
    {
      "epoch": 2.399928584181396,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 10080
    },
    {
      "epoch": 2.4001666369100754,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2798,
      "step": 10081
    },
    {
      "epoch": 2.400404689638755,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 10082
    },
    {
      "epoch": 2.4006427423674346,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 10083
    },
    {
      "epoch": 2.4008807950961137,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 10084
    },
    {
      "epoch": 2.4011188478247933,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 10085
    },
    {
      "epoch": 2.4013569005534725,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 10086
    },
    {
      "epoch": 2.401594953282152,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 10087
    },
    {
      "epoch": 2.4018330060108313,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.417,
      "step": 10088
    },
    {
      "epoch": 2.402071058739511,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 10089
    },
    {
      "epoch": 2.40230911146819,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 10090
    },
    {
      "epoch": 2.4025471641968696,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 10091
    },
    {
      "epoch": 2.4027852169255492,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 10092
    },
    {
      "epoch": 2.4030232696542284,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 10093
    },
    {
      "epoch": 2.4032613223829076,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 10094
    },
    {
      "epoch": 2.403499375111587,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 10095
    },
    {
      "epoch": 2.403737427840267,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 10096
    },
    {
      "epoch": 2.403975480568946,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 10097
    },
    {
      "epoch": 2.4042135332976255,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 10098
    },
    {
      "epoch": 2.4044515860263047,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2004,
      "step": 10099
    },
    {
      "epoch": 2.4046896387549843,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 10100
    },
    {
      "epoch": 2.4046896387549843,
      "eval_runtime": 0.5433,
      "eval_samples_per_second": 1.841,
      "eval_steps_per_second": 1.841,
      "step": 10100
    },
    {
      "epoch": 2.4049276914836635,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2944,
      "step": 10101
    },
    {
      "epoch": 2.405165744212343,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 10102
    },
    {
      "epoch": 2.4054037969410222,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 10103
    },
    {
      "epoch": 2.405641849669702,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 10104
    },
    {
      "epoch": 2.4058799023983815,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 10105
    },
    {
      "epoch": 2.4061179551270606,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 10106
    },
    {
      "epoch": 2.40635600785574,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 10107
    },
    {
      "epoch": 2.4065940605844194,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 10108
    },
    {
      "epoch": 2.406832113313099,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 10109
    },
    {
      "epoch": 2.407070166041778,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 10110
    },
    {
      "epoch": 2.4073082187704578,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 10111
    },
    {
      "epoch": 2.407546271499137,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4024,
      "step": 10112
    },
    {
      "epoch": 2.4077843242278165,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 10113
    },
    {
      "epoch": 2.4080223769564957,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 10114
    },
    {
      "epoch": 2.4082604296851753,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 10115
    },
    {
      "epoch": 2.4084984824138544,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 10116
    },
    {
      "epoch": 2.408736535142534,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 10117
    },
    {
      "epoch": 2.4089745878712137,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 10118
    },
    {
      "epoch": 2.409212640599893,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 10119
    },
    {
      "epoch": 2.4094506933285724,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 10120
    },
    {
      "epoch": 2.4096887460572516,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 10121
    },
    {
      "epoch": 2.409926798785931,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 10122
    },
    {
      "epoch": 2.4101648515146104,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 10123
    },
    {
      "epoch": 2.41040290424329,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 10124
    },
    {
      "epoch": 2.410640956971969,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 10125
    },
    {
      "epoch": 2.4108790097006487,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 10126
    },
    {
      "epoch": 2.4111170624293283,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 10127
    },
    {
      "epoch": 2.4113551151580075,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 10128
    },
    {
      "epoch": 2.411593167886687,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 10129
    },
    {
      "epoch": 2.4118312206153663,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 10130
    },
    {
      "epoch": 2.412069273344046,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 10131
    },
    {
      "epoch": 2.412307326072725,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 10132
    },
    {
      "epoch": 2.4125453788014046,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 10133
    },
    {
      "epoch": 2.412783431530084,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 10134
    },
    {
      "epoch": 2.4130214842587634,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 10135
    },
    {
      "epoch": 2.4132595369874426,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 10136
    },
    {
      "epoch": 2.413497589716122,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 10137
    },
    {
      "epoch": 2.4137356424448013,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3844,
      "step": 10138
    },
    {
      "epoch": 2.413973695173481,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 10139
    },
    {
      "epoch": 2.4142117479021605,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 10140
    },
    {
      "epoch": 2.4144498006308397,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 10141
    },
    {
      "epoch": 2.4146878533595193,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 10142
    },
    {
      "epoch": 2.4149259060881985,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 10143
    },
    {
      "epoch": 2.415163958816878,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 10144
    },
    {
      "epoch": 2.4154020115455572,
      "grad_norm": 7.59375,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 10145
    },
    {
      "epoch": 2.415640064274237,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 10146
    },
    {
      "epoch": 2.415878117002916,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 10147
    },
    {
      "epoch": 2.4161161697315956,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5843,
      "step": 10148
    },
    {
      "epoch": 2.4163542224602748,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.293,
      "step": 10149
    },
    {
      "epoch": 2.4165922751889544,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 10150
    },
    {
      "epoch": 2.4168303279176335,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 10151
    },
    {
      "epoch": 2.417068380646313,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 10152
    },
    {
      "epoch": 2.4173064333749927,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4167,
      "step": 10153
    },
    {
      "epoch": 2.417544486103672,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 10154
    },
    {
      "epoch": 2.4177825388323515,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 10155
    },
    {
      "epoch": 2.4180205915610307,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 10156
    },
    {
      "epoch": 2.4182586442897103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 10157
    },
    {
      "epoch": 2.4184966970183894,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 10158
    },
    {
      "epoch": 2.418734749747069,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 10159
    },
    {
      "epoch": 2.418972802475748,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 10160
    },
    {
      "epoch": 2.419210855204428,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 10161
    },
    {
      "epoch": 2.4194489079331074,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 10162
    },
    {
      "epoch": 2.4196869606617866,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 10163
    },
    {
      "epoch": 2.419925013390466,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 10164
    },
    {
      "epoch": 2.4201630661191453,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 10165
    },
    {
      "epoch": 2.420401118847825,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 10166
    },
    {
      "epoch": 2.420639171576504,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 10167
    },
    {
      "epoch": 2.4208772243051837,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 10168
    },
    {
      "epoch": 2.421115277033863,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 10169
    },
    {
      "epoch": 2.4213533297625425,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.5898,
      "step": 10170
    },
    {
      "epoch": 2.4215913824912216,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 10171
    },
    {
      "epoch": 2.4218294352199012,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2805,
      "step": 10172
    },
    {
      "epoch": 2.4220674879485804,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 10173
    },
    {
      "epoch": 2.42230554067726,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 10174
    },
    {
      "epoch": 2.4225435934059396,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 10175
    },
    {
      "epoch": 2.4227816461346188,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 10176
    },
    {
      "epoch": 2.4230196988632984,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.4277,
      "step": 10177
    },
    {
      "epoch": 2.4232577515919775,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 10178
    },
    {
      "epoch": 2.423495804320657,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 10179
    },
    {
      "epoch": 2.4237338570493363,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 10180
    },
    {
      "epoch": 2.423971909778016,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 10181
    },
    {
      "epoch": 2.424209962506695,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 10182
    },
    {
      "epoch": 2.4244480152353747,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 10183
    },
    {
      "epoch": 2.4246860679640543,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 10184
    },
    {
      "epoch": 2.4249241206927334,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 10185
    },
    {
      "epoch": 2.4251621734214126,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 10186
    },
    {
      "epoch": 2.425400226150092,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 10187
    },
    {
      "epoch": 2.425638278878772,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 10188
    },
    {
      "epoch": 2.425876331607451,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 10189
    },
    {
      "epoch": 2.4261143843361306,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 10190
    },
    {
      "epoch": 2.4263524370648097,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 10191
    },
    {
      "epoch": 2.4265904897934893,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 10192
    },
    {
      "epoch": 2.4268285425221685,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 10193
    },
    {
      "epoch": 2.427066595250848,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 10194
    },
    {
      "epoch": 2.4273046479795273,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 10195
    },
    {
      "epoch": 2.427542700708207,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 10196
    },
    {
      "epoch": 2.4277807534368865,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 10197
    },
    {
      "epoch": 2.4280188061655656,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 10198
    },
    {
      "epoch": 2.4282568588942453,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 10199
    },
    {
      "epoch": 2.4284949116229244,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 10200
    },
    {
      "epoch": 2.4284949116229244,
      "eval_runtime": 0.5273,
      "eval_samples_per_second": 1.896,
      "eval_steps_per_second": 1.896,
      "step": 10200
    },
    {
      "epoch": 2.428732964351604,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 10201
    },
    {
      "epoch": 2.428971017080283,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 10202
    },
    {
      "epoch": 2.429209069808963,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 10203
    },
    {
      "epoch": 2.429447122537642,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 10204
    },
    {
      "epoch": 2.4296851752663216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 10205
    },
    {
      "epoch": 2.4299232279950007,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 10206
    },
    {
      "epoch": 2.4301612807236803,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 10207
    },
    {
      "epoch": 2.4303993334523595,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4127,
      "step": 10208
    },
    {
      "epoch": 2.430637386181039,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 10209
    },
    {
      "epoch": 2.4308754389097187,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 10210
    },
    {
      "epoch": 2.431113491638398,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 10211
    },
    {
      "epoch": 2.4313515443670775,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 10212
    },
    {
      "epoch": 2.4315895970957566,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5814,
      "step": 10213
    },
    {
      "epoch": 2.431827649824436,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 10214
    },
    {
      "epoch": 2.4320657025531154,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 10215
    },
    {
      "epoch": 2.432303755281795,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 10216
    },
    {
      "epoch": 2.432541808010474,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 10217
    },
    {
      "epoch": 2.4327798607391538,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 10218
    },
    {
      "epoch": 2.4330179134678334,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 10219
    },
    {
      "epoch": 2.4332559661965125,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 10220
    },
    {
      "epoch": 2.4334940189251917,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 10221
    },
    {
      "epoch": 2.4337320716538713,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 10222
    },
    {
      "epoch": 2.433970124382551,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 10223
    },
    {
      "epoch": 2.43420817711123,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 10224
    },
    {
      "epoch": 2.4344462298399097,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 10225
    },
    {
      "epoch": 2.434684282568589,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 10226
    },
    {
      "epoch": 2.4349223352972684,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 10227
    },
    {
      "epoch": 2.4351603880259476,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4967,
      "step": 10228
    },
    {
      "epoch": 2.435398440754627,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 10229
    },
    {
      "epoch": 2.4356364934833064,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 10230
    },
    {
      "epoch": 2.435874546211986,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 10231
    },
    {
      "epoch": 2.4361125989406656,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 10232
    },
    {
      "epoch": 2.4363506516693447,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 10233
    },
    {
      "epoch": 2.4365887043980243,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 10234
    },
    {
      "epoch": 2.4368267571267035,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 10235
    },
    {
      "epoch": 2.437064809855383,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 10236
    },
    {
      "epoch": 2.4373028625840623,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2986,
      "step": 10237
    },
    {
      "epoch": 2.437540915312742,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 10238
    },
    {
      "epoch": 2.437778968041421,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 10239
    },
    {
      "epoch": 2.4380170207701006,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 10240
    },
    {
      "epoch": 2.43825507349878,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 10241
    },
    {
      "epoch": 2.4384931262274594,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1537,
      "step": 10242
    },
    {
      "epoch": 2.4387311789561386,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.5296,
      "step": 10243
    },
    {
      "epoch": 2.438969231684818,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 10244
    },
    {
      "epoch": 2.4392072844134978,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4732,
      "step": 10245
    },
    {
      "epoch": 2.439445337142177,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 10246
    },
    {
      "epoch": 2.4396833898708565,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 10247
    },
    {
      "epoch": 2.4399214425995357,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 10248
    },
    {
      "epoch": 2.4401594953282153,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 10249
    },
    {
      "epoch": 2.4403975480568945,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2559,
      "step": 10250
    },
    {
      "epoch": 2.440635600785574,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 10251
    },
    {
      "epoch": 2.4408736535142532,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 10252
    },
    {
      "epoch": 2.441111706242933,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 10253
    },
    {
      "epoch": 2.4413497589716124,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 10254
    },
    {
      "epoch": 2.4415878117002916,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 10255
    },
    {
      "epoch": 2.441825864428971,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 10256
    },
    {
      "epoch": 2.4420639171576504,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 10257
    },
    {
      "epoch": 2.44230196988633,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 10258
    },
    {
      "epoch": 2.442540022615009,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 10259
    },
    {
      "epoch": 2.4427780753436887,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 10260
    },
    {
      "epoch": 2.443016128072368,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 10261
    },
    {
      "epoch": 2.4432541808010475,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 10262
    },
    {
      "epoch": 2.4434922335297267,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 10263
    },
    {
      "epoch": 2.4437302862584063,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 10264
    },
    {
      "epoch": 2.4439683389870854,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 10265
    },
    {
      "epoch": 2.444206391715765,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 10266
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 10267
    },
    {
      "epoch": 2.444682497173124,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3984,
      "step": 10268
    },
    {
      "epoch": 2.4449205499018034,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 10269
    },
    {
      "epoch": 2.4451586026304826,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3231,
      "step": 10270
    },
    {
      "epoch": 2.445396655359162,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4361,
      "step": 10271
    },
    {
      "epoch": 2.4456347080878413,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.5991,
      "step": 10272
    },
    {
      "epoch": 2.445872760816521,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 10273
    },
    {
      "epoch": 2.4461108135452,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 10274
    },
    {
      "epoch": 2.4463488662738797,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 10275
    },
    {
      "epoch": 2.4465869190025593,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 10276
    },
    {
      "epoch": 2.4468249717312385,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 10277
    },
    {
      "epoch": 2.4470630244599176,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 10278
    },
    {
      "epoch": 2.4473010771885972,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 10279
    },
    {
      "epoch": 2.447539129917277,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 10280
    },
    {
      "epoch": 2.447777182645956,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 10281
    },
    {
      "epoch": 2.4480152353746356,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 10282
    },
    {
      "epoch": 2.4482532881033148,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 10283
    },
    {
      "epoch": 2.4484913408319944,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 10284
    },
    {
      "epoch": 2.4487293935606735,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 10285
    },
    {
      "epoch": 2.448967446289353,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 10286
    },
    {
      "epoch": 2.4492054990180323,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 10287
    },
    {
      "epoch": 2.449443551746712,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 10288
    },
    {
      "epoch": 2.4496816044753915,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3835,
      "step": 10289
    },
    {
      "epoch": 2.4499196572040707,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 10290
    },
    {
      "epoch": 2.4501577099327503,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 10291
    },
    {
      "epoch": 2.4503957626614294,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 10292
    },
    {
      "epoch": 2.450633815390109,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 10293
    },
    {
      "epoch": 2.450871868118788,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 10294
    },
    {
      "epoch": 2.451109920847468,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 10295
    },
    {
      "epoch": 2.451347973576147,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 10296
    },
    {
      "epoch": 2.4515860263048266,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 10297
    },
    {
      "epoch": 2.4518240790335057,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 10298
    },
    {
      "epoch": 2.4520621317621853,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 10299
    },
    {
      "epoch": 2.4523001844908645,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 10300
    },
    {
      "epoch": 2.4523001844908645,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 10300
    },
    {
      "epoch": 2.452538237219544,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 10301
    },
    {
      "epoch": 2.4527762899482237,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 10302
    },
    {
      "epoch": 2.453014342676903,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 10303
    },
    {
      "epoch": 2.4532523954055825,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 10304
    },
    {
      "epoch": 2.4534904481342616,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 10305
    },
    {
      "epoch": 2.4537285008629413,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 10306
    },
    {
      "epoch": 2.4539665535916204,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 10307
    },
    {
      "epoch": 2.4542046063203,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4084,
      "step": 10308
    },
    {
      "epoch": 2.454442659048979,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 10309
    },
    {
      "epoch": 2.454680711777659,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 10310
    },
    {
      "epoch": 2.4549187645063384,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4067,
      "step": 10311
    },
    {
      "epoch": 2.4551568172350176,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 10312
    },
    {
      "epoch": 2.4553948699636967,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 10313
    },
    {
      "epoch": 2.4556329226923763,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 10314
    },
    {
      "epoch": 2.455870975421056,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 10315
    },
    {
      "epoch": 2.456109028149735,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 10316
    },
    {
      "epoch": 2.4563470808784147,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 10317
    },
    {
      "epoch": 2.456585133607094,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 10318
    },
    {
      "epoch": 2.4568231863357735,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 10319
    },
    {
      "epoch": 2.4570612390644526,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.478,
      "step": 10320
    },
    {
      "epoch": 2.4572992917931322,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 10321
    },
    {
      "epoch": 2.4575373445218114,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 10322
    },
    {
      "epoch": 2.457775397250491,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 10323
    },
    {
      "epoch": 2.4580134499791706,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 10324
    },
    {
      "epoch": 2.4582515027078498,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 10325
    },
    {
      "epoch": 2.4584895554365294,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 10326
    },
    {
      "epoch": 2.4587276081652085,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 10327
    },
    {
      "epoch": 2.458965660893888,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 10328
    },
    {
      "epoch": 2.4592037136225673,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4149,
      "step": 10329
    },
    {
      "epoch": 2.459441766351247,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 10330
    },
    {
      "epoch": 2.459679819079926,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4386,
      "step": 10331
    },
    {
      "epoch": 2.4599178718086057,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 10332
    },
    {
      "epoch": 2.460155924537285,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 10333
    },
    {
      "epoch": 2.4603939772659644,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 10334
    },
    {
      "epoch": 2.4606320299946436,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 10335
    },
    {
      "epoch": 2.460870082723323,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 10336
    },
    {
      "epoch": 2.461108135452003,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4445,
      "step": 10337
    },
    {
      "epoch": 2.461346188180682,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 10338
    },
    {
      "epoch": 2.4615842409093616,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 10339
    },
    {
      "epoch": 2.4618222936380407,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 10340
    },
    {
      "epoch": 2.4620603463667203,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 10341
    },
    {
      "epoch": 2.4622983990953995,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 10342
    },
    {
      "epoch": 2.462536451824079,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 10343
    },
    {
      "epoch": 2.4627745045527583,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2783,
      "step": 10344
    },
    {
      "epoch": 2.463012557281438,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 10345
    },
    {
      "epoch": 2.4632506100101175,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 10346
    },
    {
      "epoch": 2.4634886627387966,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 10347
    },
    {
      "epoch": 2.4637267154674762,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 10348
    },
    {
      "epoch": 2.4639647681961554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 10349
    },
    {
      "epoch": 2.464202820924835,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 10350
    },
    {
      "epoch": 2.464440873653514,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 10351
    },
    {
      "epoch": 2.4646789263821938,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 10352
    },
    {
      "epoch": 2.464916979110873,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 10353
    },
    {
      "epoch": 2.4651550318395525,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 10354
    },
    {
      "epoch": 2.4653930845682317,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 10355
    },
    {
      "epoch": 2.4656311372969113,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4289,
      "step": 10356
    },
    {
      "epoch": 2.4658691900255905,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3647,
      "step": 10357
    },
    {
      "epoch": 2.46610724275427,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 10358
    },
    {
      "epoch": 2.4663452954829497,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 10359
    },
    {
      "epoch": 2.466583348211629,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 10360
    },
    {
      "epoch": 2.4668214009403084,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 10361
    },
    {
      "epoch": 2.4670594536689876,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 10362
    },
    {
      "epoch": 2.467297506397667,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 10363
    },
    {
      "epoch": 2.4675355591263464,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.5361,
      "step": 10364
    },
    {
      "epoch": 2.467773611855026,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.283,
      "step": 10365
    },
    {
      "epoch": 2.468011664583705,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 10366
    },
    {
      "epoch": 2.4682497173123847,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2748,
      "step": 10367
    },
    {
      "epoch": 2.4684877700410643,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.365,
      "step": 10368
    },
    {
      "epoch": 2.4687258227697435,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2518,
      "step": 10369
    },
    {
      "epoch": 2.4689638754984227,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 10370
    },
    {
      "epoch": 2.4692019282271023,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 10371
    },
    {
      "epoch": 2.469439980955782,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 10372
    },
    {
      "epoch": 2.469678033684461,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 10373
    },
    {
      "epoch": 2.4699160864131406,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 10374
    },
    {
      "epoch": 2.47015413914182,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 10375
    },
    {
      "epoch": 2.4703921918704994,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 10376
    },
    {
      "epoch": 2.4706302445991786,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 10377
    },
    {
      "epoch": 2.470868297327858,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 10378
    },
    {
      "epoch": 2.4711063500565373,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 10379
    },
    {
      "epoch": 2.471344402785217,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 10380
    },
    {
      "epoch": 2.4715824555138965,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 10381
    },
    {
      "epoch": 2.4718205082425757,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 10382
    },
    {
      "epoch": 2.4720585609712553,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 10383
    },
    {
      "epoch": 2.4722966136999345,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 10384
    },
    {
      "epoch": 2.472534666428614,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 10385
    },
    {
      "epoch": 2.4727727191572932,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4101,
      "step": 10386
    },
    {
      "epoch": 2.473010771885973,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 10387
    },
    {
      "epoch": 2.473248824614652,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3792,
      "step": 10388
    },
    {
      "epoch": 2.4734868773433316,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 10389
    },
    {
      "epoch": 2.4737249300720108,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 10390
    },
    {
      "epoch": 2.4739629828006904,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4279,
      "step": 10391
    },
    {
      "epoch": 2.4742010355293695,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 10392
    },
    {
      "epoch": 2.474439088258049,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 10393
    },
    {
      "epoch": 2.4746771409867288,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 10394
    },
    {
      "epoch": 2.474915193715408,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 10395
    },
    {
      "epoch": 2.4751532464440875,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 10396
    },
    {
      "epoch": 2.4753912991727667,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 10397
    },
    {
      "epoch": 2.4756293519014463,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 10398
    },
    {
      "epoch": 2.4758674046301254,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 10399
    },
    {
      "epoch": 2.476105457358805,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3272,
      "step": 10400
    },
    {
      "epoch": 2.476105457358805,
      "eval_runtime": 0.5539,
      "eval_samples_per_second": 1.805,
      "eval_steps_per_second": 1.805,
      "step": 10400
    },
    {
      "epoch": 2.476343510087484,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2716,
      "step": 10401
    },
    {
      "epoch": 2.476581562816164,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 10402
    },
    {
      "epoch": 2.4768196155448434,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 10403
    },
    {
      "epoch": 2.4770576682735226,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3426,
      "step": 10404
    },
    {
      "epoch": 2.4772957210022017,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 10405
    },
    {
      "epoch": 2.4775337737308814,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3722,
      "step": 10406
    },
    {
      "epoch": 2.477771826459561,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 10407
    },
    {
      "epoch": 2.47800987918824,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 10408
    },
    {
      "epoch": 2.4782479319169197,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 10409
    },
    {
      "epoch": 2.478485984645599,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 10410
    },
    {
      "epoch": 2.4787240373742785,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 10411
    },
    {
      "epoch": 2.4789620901029576,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 10412
    },
    {
      "epoch": 2.4792001428316373,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 10413
    },
    {
      "epoch": 2.4794381955603164,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 10414
    },
    {
      "epoch": 2.479676248288996,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 10415
    },
    {
      "epoch": 2.4799143010176756,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 10416
    },
    {
      "epoch": 2.480152353746355,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4964,
      "step": 10417
    },
    {
      "epoch": 2.4803904064750344,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 10418
    },
    {
      "epoch": 2.4806284592037136,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 10419
    },
    {
      "epoch": 2.480866511932393,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 10420
    },
    {
      "epoch": 2.4811045646610723,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 10421
    },
    {
      "epoch": 2.481342617389752,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 10422
    },
    {
      "epoch": 2.481580670118431,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 10423
    },
    {
      "epoch": 2.4818187228471107,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 10424
    },
    {
      "epoch": 2.48205677557579,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 10425
    },
    {
      "epoch": 2.4822948283044695,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3333,
      "step": 10426
    },
    {
      "epoch": 2.4825328810331486,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 10427
    },
    {
      "epoch": 2.4827709337618282,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 10428
    },
    {
      "epoch": 2.483008986490508,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3897,
      "step": 10429
    },
    {
      "epoch": 2.483247039219187,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.5166,
      "step": 10430
    },
    {
      "epoch": 2.4834850919478666,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 10431
    },
    {
      "epoch": 2.4837231446765458,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 10432
    },
    {
      "epoch": 2.4839611974052254,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4148,
      "step": 10433
    },
    {
      "epoch": 2.4841992501339045,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 10434
    },
    {
      "epoch": 2.484437302862584,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 10435
    },
    {
      "epoch": 2.4846753555912633,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 10436
    },
    {
      "epoch": 2.484913408319943,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 10437
    },
    {
      "epoch": 2.4851514610486225,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 10438
    },
    {
      "epoch": 2.4853895137773017,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 10439
    },
    {
      "epoch": 2.4856275665059813,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 10440
    },
    {
      "epoch": 2.4858656192346604,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 10441
    },
    {
      "epoch": 2.48610367196334,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 10442
    },
    {
      "epoch": 2.486341724692019,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 10443
    },
    {
      "epoch": 2.486579777420699,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 10444
    },
    {
      "epoch": 2.486817830149378,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1914,
      "step": 10445
    },
    {
      "epoch": 2.4870558828780576,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4064,
      "step": 10446
    },
    {
      "epoch": 2.4872939356067367,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3871,
      "step": 10447
    },
    {
      "epoch": 2.4875319883354163,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3265,
      "step": 10448
    },
    {
      "epoch": 2.4877700410640955,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3919,
      "step": 10449
    },
    {
      "epoch": 2.488008093792775,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 10450
    },
    {
      "epoch": 2.4882461465214547,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 10451
    },
    {
      "epoch": 2.488484199250134,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 10452
    },
    {
      "epoch": 2.4887222519788135,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 10453
    },
    {
      "epoch": 2.4889603047074926,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 10454
    },
    {
      "epoch": 2.4891983574361722,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 10455
    },
    {
      "epoch": 2.4894364101648514,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 10456
    },
    {
      "epoch": 2.489674462893531,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 10457
    },
    {
      "epoch": 2.48991251562221,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 10458
    },
    {
      "epoch": 2.4901505683508898,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 10459
    },
    {
      "epoch": 2.490388621079569,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 10460
    },
    {
      "epoch": 2.4906266738082485,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 10461
    },
    {
      "epoch": 2.4908647265369277,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 10462
    },
    {
      "epoch": 2.4911027792656073,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 10463
    },
    {
      "epoch": 2.491340831994287,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 10464
    },
    {
      "epoch": 2.491578884722966,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 10465
    },
    {
      "epoch": 2.4918169374516457,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9642,
      "step": 10466
    },
    {
      "epoch": 2.492054990180325,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 10467
    },
    {
      "epoch": 2.4922930429090044,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 10468
    },
    {
      "epoch": 2.4925310956376836,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 10469
    },
    {
      "epoch": 2.492769148366363,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2442,
      "step": 10470
    },
    {
      "epoch": 2.4930072010950424,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 10471
    },
    {
      "epoch": 2.493245253823722,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 10472
    },
    {
      "epoch": 2.4934833065524016,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 10473
    },
    {
      "epoch": 2.4937213592810807,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2977,
      "step": 10474
    },
    {
      "epoch": 2.4939594120097603,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.5545,
      "step": 10475
    },
    {
      "epoch": 2.4941974647384395,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 10476
    },
    {
      "epoch": 2.494435517467119,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 10477
    },
    {
      "epoch": 2.4946735701957983,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 10478
    },
    {
      "epoch": 2.494911622924478,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4257,
      "step": 10479
    },
    {
      "epoch": 2.495149675653157,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 10480
    },
    {
      "epoch": 2.4953877283818366,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 10481
    },
    {
      "epoch": 2.495625781110516,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 10482
    },
    {
      "epoch": 2.4958638338391954,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3464,
      "step": 10483
    },
    {
      "epoch": 2.4961018865678746,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 10484
    },
    {
      "epoch": 2.496339939296554,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 10485
    },
    {
      "epoch": 2.496577992025234,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 10486
    },
    {
      "epoch": 2.496816044753913,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 10487
    },
    {
      "epoch": 2.4970540974825925,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 10488
    },
    {
      "epoch": 2.4972921502112717,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 10489
    },
    {
      "epoch": 2.4975302029399513,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3764,
      "step": 10490
    },
    {
      "epoch": 2.4977682556686305,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 10491
    },
    {
      "epoch": 2.49800630839731,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 10492
    },
    {
      "epoch": 2.4982443611259892,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 10493
    },
    {
      "epoch": 2.498482413854669,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 10494
    },
    {
      "epoch": 2.4987204665833485,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 10495
    },
    {
      "epoch": 2.4989585193120276,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 10496
    },
    {
      "epoch": 2.4991965720407068,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 10497
    },
    {
      "epoch": 2.4994346247693864,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 10498
    },
    {
      "epoch": 2.499672677498066,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 10499
    },
    {
      "epoch": 2.499910730226745,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 10500
    },
    {
      "epoch": 2.499910730226745,
      "eval_runtime": 0.5291,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 10500
    },
    {
      "epoch": 2.5001487829554248,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4253,
      "step": 10501
    },
    {
      "epoch": 2.500386835684104,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 10502
    },
    {
      "epoch": 2.5006248884127835,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 10503
    },
    {
      "epoch": 2.5008629411414627,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 10504
    },
    {
      "epoch": 2.5011009938701423,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 10505
    },
    {
      "epoch": 2.5013390465988214,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 10506
    },
    {
      "epoch": 2.501577099327501,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4066,
      "step": 10507
    },
    {
      "epoch": 2.5018151520561807,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 10508
    },
    {
      "epoch": 2.50205320478486,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 10509
    },
    {
      "epoch": 2.502291257513539,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 10510
    },
    {
      "epoch": 2.5025293102422186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 10511
    },
    {
      "epoch": 2.502767362970898,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 10512
    },
    {
      "epoch": 2.5030054156995774,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 10513
    },
    {
      "epoch": 2.503243468428257,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 10514
    },
    {
      "epoch": 2.503481521156936,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 10515
    },
    {
      "epoch": 2.5037195738856157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 10516
    },
    {
      "epoch": 2.5039576266142953,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 10517
    },
    {
      "epoch": 2.5041956793429745,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 10518
    },
    {
      "epoch": 2.5044337320716537,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3484,
      "step": 10519
    },
    {
      "epoch": 2.5046717848003333,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.424,
      "step": 10520
    },
    {
      "epoch": 2.504909837529013,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 10521
    },
    {
      "epoch": 2.505147890257692,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 10522
    },
    {
      "epoch": 2.5053859429863716,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3762,
      "step": 10523
    },
    {
      "epoch": 2.505623995715051,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 10524
    },
    {
      "epoch": 2.5058620484437304,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 10525
    },
    {
      "epoch": 2.5061001011724096,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3544,
      "step": 10526
    },
    {
      "epoch": 2.506338153901089,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 10527
    },
    {
      "epoch": 2.5065762066297683,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 10528
    },
    {
      "epoch": 2.506814259358448,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 10529
    },
    {
      "epoch": 2.5070523120871275,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1899,
      "step": 10530
    },
    {
      "epoch": 2.5072903648158067,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 10531
    },
    {
      "epoch": 2.507528417544486,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 10532
    },
    {
      "epoch": 2.5077664702731655,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 10533
    },
    {
      "epoch": 2.508004523001845,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 10534
    },
    {
      "epoch": 2.5082425757305242,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 10535
    },
    {
      "epoch": 2.508480628459204,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 10536
    },
    {
      "epoch": 2.508718681187883,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 10537
    },
    {
      "epoch": 2.5089567339165626,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 10538
    },
    {
      "epoch": 2.5091947866452418,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 10539
    },
    {
      "epoch": 2.5094328393739214,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3205,
      "step": 10540
    },
    {
      "epoch": 2.5096708921026005,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 10541
    },
    {
      "epoch": 2.50990894483128,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 10542
    },
    {
      "epoch": 2.5101469975599597,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 10543
    },
    {
      "epoch": 2.510385050288639,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 10544
    },
    {
      "epoch": 2.510623103017318,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 10545
    },
    {
      "epoch": 2.5108611557459977,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 10546
    },
    {
      "epoch": 2.5110992084746773,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 10547
    },
    {
      "epoch": 2.5113372612033564,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 10548
    },
    {
      "epoch": 2.511575313932036,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 10549
    },
    {
      "epoch": 2.511813366660715,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 10550
    },
    {
      "epoch": 2.512051419389395,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 10551
    },
    {
      "epoch": 2.5122894721180744,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 10552
    },
    {
      "epoch": 2.5125275248467536,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2421,
      "step": 10553
    },
    {
      "epoch": 2.5127655775754327,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 10554
    },
    {
      "epoch": 2.5130036303041123,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 10555
    },
    {
      "epoch": 2.513241683032792,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 10556
    },
    {
      "epoch": 2.513479735761471,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 10557
    },
    {
      "epoch": 2.5137177884901507,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 10558
    },
    {
      "epoch": 2.51395584121883,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 10559
    },
    {
      "epoch": 2.5141938939475095,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 10560
    },
    {
      "epoch": 2.5144319466761886,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 10561
    },
    {
      "epoch": 2.5146699994048682,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 10562
    },
    {
      "epoch": 2.5149080521335474,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 10563
    },
    {
      "epoch": 2.515146104862227,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 10564
    },
    {
      "epoch": 2.5153841575909066,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 10565
    },
    {
      "epoch": 2.5156222103195858,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 10566
    },
    {
      "epoch": 2.515860263048265,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 10567
    },
    {
      "epoch": 2.5160983157769445,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3849,
      "step": 10568
    },
    {
      "epoch": 2.516336368505624,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.5414,
      "step": 10569
    },
    {
      "epoch": 2.5165744212343033,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 10570
    },
    {
      "epoch": 2.516812473962983,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.4132,
      "step": 10571
    },
    {
      "epoch": 2.517050526691662,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 10572
    },
    {
      "epoch": 2.5172885794203417,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 10573
    },
    {
      "epoch": 2.5175266321490213,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 10574
    },
    {
      "epoch": 2.5177646848777004,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 10575
    },
    {
      "epoch": 2.5180027376063796,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 10576
    },
    {
      "epoch": 2.518240790335059,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 10577
    },
    {
      "epoch": 2.518478843063739,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 10578
    },
    {
      "epoch": 2.518716895792418,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 10579
    },
    {
      "epoch": 2.5189549485210976,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 10580
    },
    {
      "epoch": 2.5191930012497767,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4703,
      "step": 10581
    },
    {
      "epoch": 2.5194310539784563,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 10582
    },
    {
      "epoch": 2.5196691067071355,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 10583
    },
    {
      "epoch": 2.519907159435815,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 10584
    },
    {
      "epoch": 2.5201452121644943,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 10585
    },
    {
      "epoch": 2.520383264893174,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 10586
    },
    {
      "epoch": 2.5206213176218535,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 10587
    },
    {
      "epoch": 2.5208593703505326,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 10588
    },
    {
      "epoch": 2.521097423079212,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 10589
    },
    {
      "epoch": 2.5213354758078914,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 10590
    },
    {
      "epoch": 2.521573528536571,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 10591
    },
    {
      "epoch": 2.52181158126525,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 10592
    },
    {
      "epoch": 2.52204963399393,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 10593
    },
    {
      "epoch": 2.522287686722609,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 10594
    },
    {
      "epoch": 2.5225257394512886,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 10595
    },
    {
      "epoch": 2.5227637921799677,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 10596
    },
    {
      "epoch": 2.5230018449086473,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 10597
    },
    {
      "epoch": 2.5232398976373265,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 10598
    },
    {
      "epoch": 2.523477950366006,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 10599
    },
    {
      "epoch": 2.5237160030946857,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 10600
    },
    {
      "epoch": 2.5237160030946857,
      "eval_runtime": 0.5138,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 10600
    },
    {
      "epoch": 2.523954055823365,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 10601
    },
    {
      "epoch": 2.524192108552044,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 10602
    },
    {
      "epoch": 2.5244301612807236,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 10603
    },
    {
      "epoch": 2.5246682140094032,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 10604
    },
    {
      "epoch": 2.5249062667380824,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 10605
    },
    {
      "epoch": 2.525144319466762,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 10606
    },
    {
      "epoch": 2.525382372195441,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 10607
    },
    {
      "epoch": 2.5256204249241208,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 10608
    },
    {
      "epoch": 2.5258584776528004,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 10609
    },
    {
      "epoch": 2.5260965303814795,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 10610
    },
    {
      "epoch": 2.5263345831101587,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 10611
    },
    {
      "epoch": 2.5265726358388383,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 10612
    },
    {
      "epoch": 2.526810688567518,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 10613
    },
    {
      "epoch": 2.527048741296197,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 10614
    },
    {
      "epoch": 2.5272867940248767,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 10615
    },
    {
      "epoch": 2.527524846753556,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 10616
    },
    {
      "epoch": 2.5277628994822354,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3307,
      "step": 10617
    },
    {
      "epoch": 2.5280009522109146,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4624,
      "step": 10618
    },
    {
      "epoch": 2.528239004939594,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 10619
    },
    {
      "epoch": 2.5284770576682734,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 10620
    },
    {
      "epoch": 2.528715110396953,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 10621
    },
    {
      "epoch": 2.5289531631256326,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 10622
    },
    {
      "epoch": 2.5291912158543117,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 10623
    },
    {
      "epoch": 2.529429268582991,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 10624
    },
    {
      "epoch": 2.5296673213116705,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 10625
    },
    {
      "epoch": 2.52990537404035,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4658,
      "step": 10626
    },
    {
      "epoch": 2.5301434267690293,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 10627
    },
    {
      "epoch": 2.530381479497709,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 10628
    },
    {
      "epoch": 2.530619532226388,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2573,
      "step": 10629
    },
    {
      "epoch": 2.5308575849550676,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 10630
    },
    {
      "epoch": 2.531095637683747,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3983,
      "step": 10631
    },
    {
      "epoch": 2.5313336904124264,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 10632
    },
    {
      "epoch": 2.5315717431411056,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4137,
      "step": 10633
    },
    {
      "epoch": 2.531809795869785,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 10634
    },
    {
      "epoch": 2.5320478485984648,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3618,
      "step": 10635
    },
    {
      "epoch": 2.532285901327144,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 10636
    },
    {
      "epoch": 2.532523954055823,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 10637
    },
    {
      "epoch": 2.5327620067845027,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 10638
    },
    {
      "epoch": 2.5330000595131823,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 10639
    },
    {
      "epoch": 2.5332381122418615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 10640
    },
    {
      "epoch": 2.533476164970541,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 10641
    },
    {
      "epoch": 2.5337142176992202,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.34,
      "step": 10642
    },
    {
      "epoch": 2.5339522704279,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 10643
    },
    {
      "epoch": 2.5341903231565794,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 10644
    },
    {
      "epoch": 2.5344283758852586,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 10645
    },
    {
      "epoch": 2.5346664286139378,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 10646
    },
    {
      "epoch": 2.5349044813426174,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 10647
    },
    {
      "epoch": 2.535142534071297,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 10648
    },
    {
      "epoch": 2.535380586799976,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 10649
    },
    {
      "epoch": 2.5356186395286557,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 10650
    },
    {
      "epoch": 2.535856692257335,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 10651
    },
    {
      "epoch": 2.5360947449860145,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 10652
    },
    {
      "epoch": 2.5363327977146937,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 10653
    },
    {
      "epoch": 2.5365708504433733,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 10654
    },
    {
      "epoch": 2.5368089031720524,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 10655
    },
    {
      "epoch": 2.537046955900732,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 10656
    },
    {
      "epoch": 2.5372850086294116,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 10657
    },
    {
      "epoch": 2.537523061358091,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 10658
    },
    {
      "epoch": 2.53776111408677,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 10659
    },
    {
      "epoch": 2.5379991668154496,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 10660
    },
    {
      "epoch": 2.538237219544129,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 10661
    },
    {
      "epoch": 2.5384752722728083,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 10662
    },
    {
      "epoch": 2.538713325001488,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4484,
      "step": 10663
    },
    {
      "epoch": 2.538951377730167,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 10664
    },
    {
      "epoch": 2.5391894304588467,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2532,
      "step": 10665
    },
    {
      "epoch": 2.5394274831875263,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 10666
    },
    {
      "epoch": 2.5396655359162055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 10667
    },
    {
      "epoch": 2.5399035886448846,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 10668
    },
    {
      "epoch": 2.5401416413735642,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 10669
    },
    {
      "epoch": 2.540379694102244,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 10670
    },
    {
      "epoch": 2.540617746830923,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 10671
    },
    {
      "epoch": 2.5408557995596026,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 10672
    },
    {
      "epoch": 2.5410938522882818,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4013,
      "step": 10673
    },
    {
      "epoch": 2.5413319050169614,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 10674
    },
    {
      "epoch": 2.5415699577456405,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 10675
    },
    {
      "epoch": 2.54180801047432,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 10676
    },
    {
      "epoch": 2.5420460632029993,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 10677
    },
    {
      "epoch": 2.542284115931679,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 10678
    },
    {
      "epoch": 2.5425221686603585,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 10679
    },
    {
      "epoch": 2.5427602213890377,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 10680
    },
    {
      "epoch": 2.542998274117717,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 10681
    },
    {
      "epoch": 2.5432363268463964,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 10682
    },
    {
      "epoch": 2.543474379575076,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 10683
    },
    {
      "epoch": 2.543712432303755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 10684
    },
    {
      "epoch": 2.543950485032435,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3236,
      "step": 10685
    },
    {
      "epoch": 2.544188537761114,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 10686
    },
    {
      "epoch": 2.5444265904897936,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 10687
    },
    {
      "epoch": 2.5446646432184727,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 10688
    },
    {
      "epoch": 2.5449026959471523,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 10689
    },
    {
      "epoch": 2.5451407486758315,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 10690
    },
    {
      "epoch": 2.545378801404511,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 10691
    },
    {
      "epoch": 2.5456168541331907,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4601,
      "step": 10692
    },
    {
      "epoch": 2.54585490686187,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 10693
    },
    {
      "epoch": 2.546092959590549,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 10694
    },
    {
      "epoch": 2.5463310123192286,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 10695
    },
    {
      "epoch": 2.5465690650479083,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 10696
    },
    {
      "epoch": 2.5468071177765874,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 10697
    },
    {
      "epoch": 2.547045170505267,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 10698
    },
    {
      "epoch": 2.547283223233946,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5044,
      "step": 10699
    },
    {
      "epoch": 2.547521275962626,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 10700
    },
    {
      "epoch": 2.547521275962626,
      "eval_runtime": 0.5226,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 10700
    },
    {
      "epoch": 2.5477593286913054,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.7065,
      "step": 10701
    },
    {
      "epoch": 2.5479973814199846,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 10702
    },
    {
      "epoch": 2.5482354341486637,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 10703
    },
    {
      "epoch": 2.5484734868773433,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 10704
    },
    {
      "epoch": 2.548711539606023,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 10705
    },
    {
      "epoch": 2.548949592334702,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 10706
    },
    {
      "epoch": 2.5491876450633817,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 10707
    },
    {
      "epoch": 2.549425697792061,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 2.2132,
      "step": 10708
    },
    {
      "epoch": 2.5496637505207405,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 10709
    },
    {
      "epoch": 2.5499018032494196,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3875,
      "step": 10710
    },
    {
      "epoch": 2.5501398559780992,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 10711
    },
    {
      "epoch": 2.5503779087067784,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 10712
    },
    {
      "epoch": 2.550615961435458,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7931,
      "step": 10713
    },
    {
      "epoch": 2.5508540141641376,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 10714
    },
    {
      "epoch": 2.5510920668928168,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 10715
    },
    {
      "epoch": 2.551330119621496,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 10716
    },
    {
      "epoch": 2.5515681723501755,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 10717
    },
    {
      "epoch": 2.551806225078855,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 10718
    },
    {
      "epoch": 2.5520442778075343,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 10719
    },
    {
      "epoch": 2.552282330536214,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 10720
    },
    {
      "epoch": 2.552520383264893,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 10721
    },
    {
      "epoch": 2.5527584359935727,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 10722
    },
    {
      "epoch": 2.552996488722252,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 10723
    },
    {
      "epoch": 2.5532345414509314,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 10724
    },
    {
      "epoch": 2.5534725941796106,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 10725
    },
    {
      "epoch": 2.55371064690829,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 10726
    },
    {
      "epoch": 2.55394869963697,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 10727
    },
    {
      "epoch": 2.554186752365649,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5268,
      "step": 10728
    },
    {
      "epoch": 2.554424805094328,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 10729
    },
    {
      "epoch": 2.5546628578230077,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 10730
    },
    {
      "epoch": 2.5549009105516873,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 10731
    },
    {
      "epoch": 2.5551389632803665,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 10732
    },
    {
      "epoch": 2.555377016009046,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 10733
    },
    {
      "epoch": 2.5556150687377253,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 10734
    },
    {
      "epoch": 2.555853121466405,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2649,
      "step": 10735
    },
    {
      "epoch": 2.5560911741950845,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 10736
    },
    {
      "epoch": 2.5563292269237636,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 10737
    },
    {
      "epoch": 2.556567279652443,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 10738
    },
    {
      "epoch": 2.5568053323811224,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 10739
    },
    {
      "epoch": 2.557043385109802,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 10740
    },
    {
      "epoch": 2.557281437838481,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 10741
    },
    {
      "epoch": 2.5575194905671608,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 10742
    },
    {
      "epoch": 2.55775754329584,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 10743
    },
    {
      "epoch": 2.5579955960245195,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 10744
    },
    {
      "epoch": 2.5582336487531987,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4073,
      "step": 10745
    },
    {
      "epoch": 2.5584717014818783,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 10746
    },
    {
      "epoch": 2.5587097542105575,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 10747
    },
    {
      "epoch": 2.558947806939237,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3751,
      "step": 10748
    },
    {
      "epoch": 2.5591858596679167,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 10749
    },
    {
      "epoch": 2.559423912396596,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4177,
      "step": 10750
    },
    {
      "epoch": 2.559661965125275,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 10751
    },
    {
      "epoch": 2.5599000178539546,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 10752
    },
    {
      "epoch": 2.560138070582634,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.5975,
      "step": 10753
    },
    {
      "epoch": 2.5603761233113134,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 10754
    },
    {
      "epoch": 2.560614176039993,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 10755
    },
    {
      "epoch": 2.560852228768672,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.457,
      "step": 10756
    },
    {
      "epoch": 2.5610902814973517,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 10757
    },
    {
      "epoch": 2.561328334226031,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 10758
    },
    {
      "epoch": 2.5615663869547105,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 10759
    },
    {
      "epoch": 2.5618044396833897,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3679,
      "step": 10760
    },
    {
      "epoch": 2.5620424924120693,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2528,
      "step": 10761
    },
    {
      "epoch": 2.562280545140749,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 10762
    },
    {
      "epoch": 2.562518597869428,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 10763
    },
    {
      "epoch": 2.5627566505981076,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 10764
    },
    {
      "epoch": 2.562994703326787,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 10765
    },
    {
      "epoch": 2.5632327560554664,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 10766
    },
    {
      "epoch": 2.5634708087841456,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 10767
    },
    {
      "epoch": 2.563708861512825,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 10768
    },
    {
      "epoch": 2.5639469142415043,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4419,
      "step": 10769
    },
    {
      "epoch": 2.564184966970184,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 10770
    },
    {
      "epoch": 2.5644230196988635,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 10771
    },
    {
      "epoch": 2.5646610724275427,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 10772
    },
    {
      "epoch": 2.564899125156222,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 10773
    },
    {
      "epoch": 2.5651371778849015,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 10774
    },
    {
      "epoch": 2.565375230613581,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 10775
    },
    {
      "epoch": 2.5656132833422602,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 10776
    },
    {
      "epoch": 2.56585133607094,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 10777
    },
    {
      "epoch": 2.566089388799619,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 10778
    },
    {
      "epoch": 2.5663274415282986,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 10779
    },
    {
      "epoch": 2.5665654942569778,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 10780
    },
    {
      "epoch": 2.5668035469856574,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 10781
    },
    {
      "epoch": 2.5670415997143365,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 10782
    },
    {
      "epoch": 2.567279652443016,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 10783
    },
    {
      "epoch": 2.5675177051716958,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 10784
    },
    {
      "epoch": 2.567755757900375,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 10785
    },
    {
      "epoch": 2.567993810629054,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8155,
      "step": 10786
    },
    {
      "epoch": 2.5682318633577337,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 10787
    },
    {
      "epoch": 2.5684699160864133,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 10788
    },
    {
      "epoch": 2.5687079688150924,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 10789
    },
    {
      "epoch": 2.568946021543772,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3021,
      "step": 10790
    },
    {
      "epoch": 2.569184074272451,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4673,
      "step": 10791
    },
    {
      "epoch": 2.569422127001131,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.5573,
      "step": 10792
    },
    {
      "epoch": 2.5696601797298104,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 10793
    },
    {
      "epoch": 2.5698982324584896,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 10794
    },
    {
      "epoch": 2.5701362851871687,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 10795
    },
    {
      "epoch": 2.5703743379158484,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3348,
      "step": 10796
    },
    {
      "epoch": 2.570612390644528,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 10797
    },
    {
      "epoch": 2.570850443373207,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 10798
    },
    {
      "epoch": 2.5710884961018867,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 10799
    },
    {
      "epoch": 2.571326548830566,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 10800
    },
    {
      "epoch": 2.571326548830566,
      "eval_runtime": 0.5173,
      "eval_samples_per_second": 1.933,
      "eval_steps_per_second": 1.933,
      "step": 10800
    },
    {
      "epoch": 2.5715646015592455,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 10801
    },
    {
      "epoch": 2.5718026542879246,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 10802
    },
    {
      "epoch": 2.5720407070166043,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 10803
    },
    {
      "epoch": 2.5722787597452834,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 10804
    },
    {
      "epoch": 2.572516812473963,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3847,
      "step": 10805
    },
    {
      "epoch": 2.5727548652026426,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 10806
    },
    {
      "epoch": 2.572992917931322,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 10807
    },
    {
      "epoch": 2.573230970660001,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 10808
    },
    {
      "epoch": 2.5734690233886806,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 10809
    },
    {
      "epoch": 2.57370707611736,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 10810
    },
    {
      "epoch": 2.5739451288460393,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 10811
    },
    {
      "epoch": 2.574183181574719,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 10812
    },
    {
      "epoch": 2.574421234303398,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.5064,
      "step": 10813
    },
    {
      "epoch": 2.5746592870320777,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 10814
    },
    {
      "epoch": 2.574897339760757,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 10815
    },
    {
      "epoch": 2.5751353924894365,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 10816
    },
    {
      "epoch": 2.5753734452181156,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 10817
    },
    {
      "epoch": 2.5756114979467952,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 10818
    },
    {
      "epoch": 2.575849550675475,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 10819
    },
    {
      "epoch": 2.576087603404154,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 10820
    },
    {
      "epoch": 2.576325656132833,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 10821
    },
    {
      "epoch": 2.5765637088615128,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 10822
    },
    {
      "epoch": 2.5768017615901924,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 10823
    },
    {
      "epoch": 2.5770398143188715,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 10824
    },
    {
      "epoch": 2.577277867047551,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 10825
    },
    {
      "epoch": 2.5775159197762303,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.5886,
      "step": 10826
    },
    {
      "epoch": 2.57775397250491,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4054,
      "step": 10827
    },
    {
      "epoch": 2.5779920252335895,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 10828
    },
    {
      "epoch": 2.5782300779622687,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 10829
    },
    {
      "epoch": 2.578468130690948,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.15,
      "step": 10830
    },
    {
      "epoch": 2.5787061834196274,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 10831
    },
    {
      "epoch": 2.578944236148307,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 10832
    },
    {
      "epoch": 2.579182288876986,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 10833
    },
    {
      "epoch": 2.579420341605666,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3614,
      "step": 10834
    },
    {
      "epoch": 2.579658394334345,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 10835
    },
    {
      "epoch": 2.5798964470630246,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 10836
    },
    {
      "epoch": 2.5801344997917037,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 10837
    },
    {
      "epoch": 2.5803725525203833,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 10838
    },
    {
      "epoch": 2.5806106052490625,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 10839
    },
    {
      "epoch": 2.580848657977742,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 10840
    },
    {
      "epoch": 2.5810867107064217,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 10841
    },
    {
      "epoch": 2.581324763435101,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 10842
    },
    {
      "epoch": 2.58156281616378,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 10843
    },
    {
      "epoch": 2.5818008688924596,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 10844
    },
    {
      "epoch": 2.5820389216211392,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 10845
    },
    {
      "epoch": 2.5822769743498184,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 10846
    },
    {
      "epoch": 2.582515027078498,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 10847
    },
    {
      "epoch": 2.582753079807177,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 10848
    },
    {
      "epoch": 2.5829911325358568,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 10849
    },
    {
      "epoch": 2.583229185264536,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 10850
    },
    {
      "epoch": 2.5834672379932155,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 10851
    },
    {
      "epoch": 2.5837052907218947,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 10852
    },
    {
      "epoch": 2.5839433434505743,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 10853
    },
    {
      "epoch": 2.584181396179254,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 10854
    },
    {
      "epoch": 2.584419448907933,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 10855
    },
    {
      "epoch": 2.5846575016366127,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 10856
    },
    {
      "epoch": 2.584895554365292,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 10857
    },
    {
      "epoch": 2.5851336070939714,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 10858
    },
    {
      "epoch": 2.5853716598226506,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 10859
    },
    {
      "epoch": 2.58560971255133,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 10860
    },
    {
      "epoch": 2.5858477652800094,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 10861
    },
    {
      "epoch": 2.586085818008689,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 10862
    },
    {
      "epoch": 2.5863238707373686,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 10863
    },
    {
      "epoch": 2.5865619234660477,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 10864
    },
    {
      "epoch": 2.586799976194727,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3791,
      "step": 10865
    },
    {
      "epoch": 2.5870380289234065,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 10866
    },
    {
      "epoch": 2.587276081652086,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 10867
    },
    {
      "epoch": 2.5875141343807653,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.377,
      "step": 10868
    },
    {
      "epoch": 2.587752187109445,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 10869
    },
    {
      "epoch": 2.587990239838124,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 10870
    },
    {
      "epoch": 2.5882282925668036,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 10871
    },
    {
      "epoch": 2.588466345295483,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 10872
    },
    {
      "epoch": 2.5887043980241624,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 10873
    },
    {
      "epoch": 2.5889424507528416,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 10874
    },
    {
      "epoch": 2.589180503481521,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 10875
    },
    {
      "epoch": 2.589418556210201,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 10876
    },
    {
      "epoch": 2.58965660893888,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.909,
      "step": 10877
    },
    {
      "epoch": 2.589894661667559,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 10878
    },
    {
      "epoch": 2.5901327143962387,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5928,
      "step": 10879
    },
    {
      "epoch": 2.5903707671249183,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 10880
    },
    {
      "epoch": 2.5906088198535975,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 10881
    },
    {
      "epoch": 2.590846872582277,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 10882
    },
    {
      "epoch": 2.5910849253109562,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 10883
    },
    {
      "epoch": 2.591322978039636,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 10884
    },
    {
      "epoch": 2.5915610307683155,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3123,
      "step": 10885
    },
    {
      "epoch": 2.5917990834969946,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 10886
    },
    {
      "epoch": 2.5920371362256738,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 10887
    },
    {
      "epoch": 2.5922751889543534,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 10888
    },
    {
      "epoch": 2.592513241683033,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 10889
    },
    {
      "epoch": 2.592751294411712,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 10890
    },
    {
      "epoch": 2.5929893471403918,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 10891
    },
    {
      "epoch": 2.593227399869071,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 10892
    },
    {
      "epoch": 2.5934654525977505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 10893
    },
    {
      "epoch": 2.5937035053264297,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 10894
    },
    {
      "epoch": 2.5939415580551093,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 10895
    },
    {
      "epoch": 2.5941796107837884,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 10896
    },
    {
      "epoch": 2.594417663512468,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 10897
    },
    {
      "epoch": 2.5946557162411477,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 10898
    },
    {
      "epoch": 2.594893768969827,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 10899
    },
    {
      "epoch": 2.595131821698506,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 10900
    },
    {
      "epoch": 2.595131821698506,
      "eval_runtime": 0.5224,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 10900
    },
    {
      "epoch": 2.5953698744271856,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 10901
    },
    {
      "epoch": 2.595607927155865,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 10902
    },
    {
      "epoch": 2.5958459798845444,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3603,
      "step": 10903
    },
    {
      "epoch": 2.596084032613224,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 10904
    },
    {
      "epoch": 2.596322085341903,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 10905
    },
    {
      "epoch": 2.5965601380705827,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 10906
    },
    {
      "epoch": 2.596798190799262,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 10907
    },
    {
      "epoch": 2.5970362435279415,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 10908
    },
    {
      "epoch": 2.5972742962566207,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.237,
      "step": 10909
    },
    {
      "epoch": 2.5975123489853003,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 10910
    },
    {
      "epoch": 2.59775040171398,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2438,
      "step": 10911
    },
    {
      "epoch": 2.597988454442659,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 10912
    },
    {
      "epoch": 2.598226507171338,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 10913
    },
    {
      "epoch": 2.598464559900018,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 10914
    },
    {
      "epoch": 2.5987026126286974,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 10915
    },
    {
      "epoch": 2.5989406653573766,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4706,
      "step": 10916
    },
    {
      "epoch": 2.599178718086056,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 10917
    },
    {
      "epoch": 2.5994167708147353,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 10918
    },
    {
      "epoch": 2.599654823543415,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 10919
    },
    {
      "epoch": 2.5998928762720945,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 10920
    },
    {
      "epoch": 2.6001309290007737,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 10921
    },
    {
      "epoch": 2.600368981729453,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 10922
    },
    {
      "epoch": 2.6006070344581325,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 10923
    },
    {
      "epoch": 2.600845087186812,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 10924
    },
    {
      "epoch": 2.6010831399154912,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.558,
      "step": 10925
    },
    {
      "epoch": 2.601321192644171,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 10926
    },
    {
      "epoch": 2.60155924537285,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 10927
    },
    {
      "epoch": 2.6017972981015296,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8264,
      "step": 10928
    },
    {
      "epoch": 2.6020353508302088,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 10929
    },
    {
      "epoch": 2.6022734035588884,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 10930
    },
    {
      "epoch": 2.6025114562875675,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 10931
    },
    {
      "epoch": 2.602749509016247,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 10932
    },
    {
      "epoch": 2.6029875617449267,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 10933
    },
    {
      "epoch": 2.603225614473606,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 10934
    },
    {
      "epoch": 2.603463667202285,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 10935
    },
    {
      "epoch": 2.6037017199309647,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 10936
    },
    {
      "epoch": 2.6039397726596443,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 10937
    },
    {
      "epoch": 2.6041778253883234,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 10938
    },
    {
      "epoch": 2.604415878117003,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 10939
    },
    {
      "epoch": 2.604653930845682,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 10940
    },
    {
      "epoch": 2.604891983574362,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 10941
    },
    {
      "epoch": 2.605130036303041,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4918,
      "step": 10942
    },
    {
      "epoch": 2.6053680890317206,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4533,
      "step": 10943
    },
    {
      "epoch": 2.6056061417603997,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 10944
    },
    {
      "epoch": 2.6058441944890793,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4152,
      "step": 10945
    },
    {
      "epoch": 2.606082247217759,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 10946
    },
    {
      "epoch": 2.606320299946438,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 10947
    },
    {
      "epoch": 2.6065583526751173,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 10948
    },
    {
      "epoch": 2.606796405403797,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 10949
    },
    {
      "epoch": 2.6070344581324765,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4564,
      "step": 10950
    },
    {
      "epoch": 2.6072725108611556,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3378,
      "step": 10951
    },
    {
      "epoch": 2.6075105635898352,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 10952
    },
    {
      "epoch": 2.6077486163185144,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 10953
    },
    {
      "epoch": 2.607986669047194,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 10954
    },
    {
      "epoch": 2.6082247217758736,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3006,
      "step": 10955
    },
    {
      "epoch": 2.6084627745045528,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 10956
    },
    {
      "epoch": 2.608700827233232,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 10957
    },
    {
      "epoch": 2.6089388799619115,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 10958
    },
    {
      "epoch": 2.609176932690591,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 10959
    },
    {
      "epoch": 2.6094149854192703,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 10960
    },
    {
      "epoch": 2.60965303814795,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 10961
    },
    {
      "epoch": 2.609891090876629,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3002,
      "step": 10962
    },
    {
      "epoch": 2.6101291436053087,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3044,
      "step": 10963
    },
    {
      "epoch": 2.610367196333988,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 10964
    },
    {
      "epoch": 2.6106052490626674,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 10965
    },
    {
      "epoch": 2.6108433017913466,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 10966
    },
    {
      "epoch": 2.611081354520026,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 10967
    },
    {
      "epoch": 2.611319407248706,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 10968
    },
    {
      "epoch": 2.611557459977385,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3941,
      "step": 10969
    },
    {
      "epoch": 2.611795512706064,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 10970
    },
    {
      "epoch": 2.6120335654347437,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 10971
    },
    {
      "epoch": 2.6122716181634233,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 10972
    },
    {
      "epoch": 2.6125096708921025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 10973
    },
    {
      "epoch": 2.612747723620782,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 10974
    },
    {
      "epoch": 2.6129857763494613,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 10975
    },
    {
      "epoch": 2.613223829078141,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 10976
    },
    {
      "epoch": 2.6134618818068205,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 10977
    },
    {
      "epoch": 2.6136999345354996,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 10978
    },
    {
      "epoch": 2.613937987264179,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 10979
    },
    {
      "epoch": 2.6141760399928584,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 10980
    },
    {
      "epoch": 2.614414092721538,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 10981
    },
    {
      "epoch": 2.614652145450217,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 10982
    },
    {
      "epoch": 2.614890198178897,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1923,
      "step": 10983
    },
    {
      "epoch": 2.615128250907576,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 10984
    },
    {
      "epoch": 2.6153663036362556,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4442,
      "step": 10985
    },
    {
      "epoch": 2.6156043563649347,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 10986
    },
    {
      "epoch": 2.6158424090936143,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 10987
    },
    {
      "epoch": 2.6160804618222935,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 10988
    },
    {
      "epoch": 2.616318514550973,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 10989
    },
    {
      "epoch": 2.6165565672796527,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 10990
    },
    {
      "epoch": 2.616794620008332,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 10991
    },
    {
      "epoch": 2.617032672737011,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 10992
    },
    {
      "epoch": 2.6172707254656906,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 10993
    },
    {
      "epoch": 2.6175087781943702,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 10994
    },
    {
      "epoch": 2.6177468309230494,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 10995
    },
    {
      "epoch": 2.617984883651729,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 10996
    },
    {
      "epoch": 2.618222936380408,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 10997
    },
    {
      "epoch": 2.6184609891090878,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 10998
    },
    {
      "epoch": 2.618699041837767,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 10999
    },
    {
      "epoch": 2.6189370945664465,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 11000
    },
    {
      "epoch": 2.6189370945664465,
      "eval_runtime": 0.5146,
      "eval_samples_per_second": 1.943,
      "eval_steps_per_second": 1.943,
      "step": 11000
    },
    {
      "epoch": 2.6191751472951257,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.5524,
      "step": 11001
    },
    {
      "epoch": 2.6194132000238053,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9363,
      "step": 11002
    },
    {
      "epoch": 2.619651252752485,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 11003
    },
    {
      "epoch": 2.619889305481164,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 11004
    },
    {
      "epoch": 2.620127358209843,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.4198,
      "step": 11005
    },
    {
      "epoch": 2.620365410938523,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 11006
    },
    {
      "epoch": 2.6206034636672024,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 11007
    },
    {
      "epoch": 2.6208415163958816,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 11008
    },
    {
      "epoch": 2.621079569124561,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 11009
    },
    {
      "epoch": 2.6213176218532404,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.441,
      "step": 11010
    },
    {
      "epoch": 2.62155567458192,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 11011
    },
    {
      "epoch": 2.6217937273105996,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 11012
    },
    {
      "epoch": 2.6220317800392787,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 11013
    },
    {
      "epoch": 2.622269832767958,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 11014
    },
    {
      "epoch": 2.6225078854966375,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 11015
    },
    {
      "epoch": 2.622745938225317,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 11016
    },
    {
      "epoch": 2.6229839909539963,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 11017
    },
    {
      "epoch": 2.623222043682676,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 11018
    },
    {
      "epoch": 2.623460096411355,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 11019
    },
    {
      "epoch": 2.6236981491400346,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 11020
    },
    {
      "epoch": 2.623936201868714,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 11021
    },
    {
      "epoch": 2.6241742545973934,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 11022
    },
    {
      "epoch": 2.6244123073260726,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.5413,
      "step": 11023
    },
    {
      "epoch": 2.624650360054752,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 11024
    },
    {
      "epoch": 2.6248884127834318,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 11025
    },
    {
      "epoch": 2.625126465512111,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 11026
    },
    {
      "epoch": 2.62536451824079,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 11027
    },
    {
      "epoch": 2.6256025709694697,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 11028
    },
    {
      "epoch": 2.6258406236981493,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 11029
    },
    {
      "epoch": 2.6260786764268285,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 11030
    },
    {
      "epoch": 2.626316729155508,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 11031
    },
    {
      "epoch": 2.6265547818841872,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4396,
      "step": 11032
    },
    {
      "epoch": 2.626792834612867,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 11033
    },
    {
      "epoch": 2.627030887341546,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 11034
    },
    {
      "epoch": 2.6272689400702256,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 11035
    },
    {
      "epoch": 2.6275069927989048,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 11036
    },
    {
      "epoch": 2.6277450455275844,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 11037
    },
    {
      "epoch": 2.627983098256264,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 11038
    },
    {
      "epoch": 2.628221150984943,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 11039
    },
    {
      "epoch": 2.6284592037136223,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 11040
    },
    {
      "epoch": 2.628697256442302,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 11041
    },
    {
      "epoch": 2.6289353091709815,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2037,
      "step": 11042
    },
    {
      "epoch": 2.6291733618996607,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 11043
    },
    {
      "epoch": 2.6294114146283403,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 11044
    },
    {
      "epoch": 2.6296494673570194,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 11045
    },
    {
      "epoch": 2.629887520085699,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 11046
    },
    {
      "epoch": 2.6301255728143786,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 11047
    },
    {
      "epoch": 2.630363625543058,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.4728,
      "step": 11048
    },
    {
      "epoch": 2.630601678271737,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 11049
    },
    {
      "epoch": 2.6308397310004166,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3529,
      "step": 11050
    },
    {
      "epoch": 2.631077783729096,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 11051
    },
    {
      "epoch": 2.6313158364577753,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 11052
    },
    {
      "epoch": 2.631553889186455,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 11053
    },
    {
      "epoch": 2.631791941915134,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 2.4018,
      "step": 11054
    },
    {
      "epoch": 2.6320299946438137,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 11055
    },
    {
      "epoch": 2.632268047372493,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3602,
      "step": 11056
    },
    {
      "epoch": 2.6325061001011725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 11057
    },
    {
      "epoch": 2.6327441528298516,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 11058
    },
    {
      "epoch": 2.6329822055585312,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 11059
    },
    {
      "epoch": 2.633220258287211,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 11060
    },
    {
      "epoch": 2.63345831101589,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 11061
    },
    {
      "epoch": 2.633696363744569,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 11062
    },
    {
      "epoch": 2.6339344164732488,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 11063
    },
    {
      "epoch": 2.6341724692019284,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 11064
    },
    {
      "epoch": 2.6344105219306075,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 11065
    },
    {
      "epoch": 2.634648574659287,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 11066
    },
    {
      "epoch": 2.6348866273879663,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 11067
    },
    {
      "epoch": 2.635124680116646,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 11068
    },
    {
      "epoch": 2.635362732845325,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 11069
    },
    {
      "epoch": 2.6356007855740047,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 11070
    },
    {
      "epoch": 2.635838838302684,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 11071
    },
    {
      "epoch": 2.6360768910313634,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 11072
    },
    {
      "epoch": 2.636314943760043,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 11073
    },
    {
      "epoch": 2.636552996488722,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 11074
    },
    {
      "epoch": 2.636791049217402,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 11075
    },
    {
      "epoch": 2.637029101946081,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 11076
    },
    {
      "epoch": 2.6372671546747606,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 11077
    },
    {
      "epoch": 2.6375052074034397,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 11078
    },
    {
      "epoch": 2.6377432601321193,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 11079
    },
    {
      "epoch": 2.6379813128607985,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 11080
    },
    {
      "epoch": 2.638219365589478,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 11081
    },
    {
      "epoch": 2.6384574183181577,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 11082
    },
    {
      "epoch": 2.638695471046837,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 11083
    },
    {
      "epoch": 2.638933523775516,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 11084
    },
    {
      "epoch": 2.6391715765041956,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 11085
    },
    {
      "epoch": 2.6394096292328753,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 11086
    },
    {
      "epoch": 2.6396476819615544,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 11087
    },
    {
      "epoch": 2.639885734690234,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 11088
    },
    {
      "epoch": 2.640123787418913,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 11089
    },
    {
      "epoch": 2.640361840147593,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 11090
    },
    {
      "epoch": 2.640599892876272,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1368,
      "step": 11091
    },
    {
      "epoch": 2.6408379456049516,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 11092
    },
    {
      "epoch": 2.6410759983336307,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.408,
      "step": 11093
    },
    {
      "epoch": 2.6413140510623103,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 11094
    },
    {
      "epoch": 2.64155210379099,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 11095
    },
    {
      "epoch": 2.641790156519669,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8374,
      "step": 11096
    },
    {
      "epoch": 2.6420282092483482,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 11097
    },
    {
      "epoch": 2.642266261977028,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 11098
    },
    {
      "epoch": 2.6425043147057075,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 11099
    },
    {
      "epoch": 2.6427423674343866,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 11100
    },
    {
      "epoch": 2.6427423674343866,
      "eval_runtime": 0.5299,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 11100
    },
    {
      "epoch": 2.6429804201630662,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 11101
    },
    {
      "epoch": 2.6432184728917454,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 11102
    },
    {
      "epoch": 2.643456525620425,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 11103
    },
    {
      "epoch": 2.6436945783491046,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4043,
      "step": 11104
    },
    {
      "epoch": 2.6439326310777838,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 11105
    },
    {
      "epoch": 2.644170683806463,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 11106
    },
    {
      "epoch": 2.6444087365351425,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 11107
    },
    {
      "epoch": 2.644646789263822,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 11108
    },
    {
      "epoch": 2.6448848419925013,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3892,
      "step": 11109
    },
    {
      "epoch": 2.645122894721181,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 11110
    },
    {
      "epoch": 2.64536094744986,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 11111
    },
    {
      "epoch": 2.6455990001785397,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3467,
      "step": 11112
    },
    {
      "epoch": 2.645837052907219,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 11113
    },
    {
      "epoch": 2.6460751056358984,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 11114
    },
    {
      "epoch": 2.6463131583645776,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5373,
      "step": 11115
    },
    {
      "epoch": 2.646551211093257,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7992,
      "step": 11116
    },
    {
      "epoch": 2.646789263821937,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 11117
    },
    {
      "epoch": 2.647027316550616,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 11118
    },
    {
      "epoch": 2.647265369279295,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 11119
    },
    {
      "epoch": 2.6475034220079747,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 11120
    },
    {
      "epoch": 2.6477414747366543,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 11121
    },
    {
      "epoch": 2.6479795274653335,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 11122
    },
    {
      "epoch": 2.648217580194013,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.5496,
      "step": 11123
    },
    {
      "epoch": 2.6484556329226923,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 11124
    },
    {
      "epoch": 2.648693685651372,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 11125
    },
    {
      "epoch": 2.648931738380051,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 11126
    },
    {
      "epoch": 2.6491697911087306,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 11127
    },
    {
      "epoch": 2.64940784383741,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 11128
    },
    {
      "epoch": 2.6496458965660894,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 11129
    },
    {
      "epoch": 2.649883949294769,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 11130
    },
    {
      "epoch": 2.650122002023448,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 11131
    },
    {
      "epoch": 2.6503600547521273,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 11132
    },
    {
      "epoch": 2.650598107480807,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5026,
      "step": 11133
    },
    {
      "epoch": 2.6508361602094865,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 11134
    },
    {
      "epoch": 2.6510742129381657,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2283,
      "step": 11135
    },
    {
      "epoch": 2.6513122656668453,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 11136
    },
    {
      "epoch": 2.6515503183955245,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 11137
    },
    {
      "epoch": 2.651788371124204,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 11138
    },
    {
      "epoch": 2.6520264238528837,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 11139
    },
    {
      "epoch": 2.652264476581563,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 11140
    },
    {
      "epoch": 2.652502529310242,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4618,
      "step": 11141
    },
    {
      "epoch": 2.6527405820389216,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 11142
    },
    {
      "epoch": 2.652978634767601,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 11143
    },
    {
      "epoch": 2.6532166874962804,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 11144
    },
    {
      "epoch": 2.65345474022496,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3139,
      "step": 11145
    },
    {
      "epoch": 2.653692792953639,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 11146
    },
    {
      "epoch": 2.6539308456823187,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4071,
      "step": 11147
    },
    {
      "epoch": 2.654168898410998,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 11148
    },
    {
      "epoch": 2.6544069511396775,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 11149
    },
    {
      "epoch": 2.6546450038683567,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 11150
    },
    {
      "epoch": 2.6548830565970363,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 11151
    },
    {
      "epoch": 2.655121109325716,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 11152
    },
    {
      "epoch": 2.655359162054395,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 11153
    },
    {
      "epoch": 2.655597214783074,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 11154
    },
    {
      "epoch": 2.655835267511754,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 11155
    },
    {
      "epoch": 2.6560733202404334,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 11156
    },
    {
      "epoch": 2.6563113729691126,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 11157
    },
    {
      "epoch": 2.656549425697792,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 11158
    },
    {
      "epoch": 2.6567874784264713,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4686,
      "step": 11159
    },
    {
      "epoch": 2.657025531155151,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 11160
    },
    {
      "epoch": 2.65726358388383,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 11161
    },
    {
      "epoch": 2.6575016366125097,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 11162
    },
    {
      "epoch": 2.657739689341189,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 11163
    },
    {
      "epoch": 2.6579777420698685,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.4533,
      "step": 11164
    },
    {
      "epoch": 2.658215794798548,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 11165
    },
    {
      "epoch": 2.6584538475272272,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 11166
    },
    {
      "epoch": 2.658691900255907,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 11167
    },
    {
      "epoch": 2.658929952984586,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 11168
    },
    {
      "epoch": 2.6591680057132656,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 11169
    },
    {
      "epoch": 2.6594060584419448,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 11170
    },
    {
      "epoch": 2.6596441111706244,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 11171
    },
    {
      "epoch": 2.6598821638993035,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 11172
    },
    {
      "epoch": 2.660120216627983,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 11173
    },
    {
      "epoch": 2.6603582693566628,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.544,
      "step": 11174
    },
    {
      "epoch": 2.660596322085342,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 11175
    },
    {
      "epoch": 2.660834374814021,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 11176
    },
    {
      "epoch": 2.6610724275427007,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4927,
      "step": 11177
    },
    {
      "epoch": 2.6613104802713803,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 11178
    },
    {
      "epoch": 2.6615485330000594,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 11179
    },
    {
      "epoch": 2.661786585728739,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5099,
      "step": 11180
    },
    {
      "epoch": 2.662024638457418,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.414,
      "step": 11181
    },
    {
      "epoch": 2.662262691186098,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 11182
    },
    {
      "epoch": 2.662500743914777,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 11183
    },
    {
      "epoch": 2.6627387966434566,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 11184
    },
    {
      "epoch": 2.6629768493721357,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 11185
    },
    {
      "epoch": 2.6632149021008154,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 11186
    },
    {
      "epoch": 2.663452954829495,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 11187
    },
    {
      "epoch": 2.663691007558174,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 11188
    },
    {
      "epoch": 2.6639290602868533,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 11189
    },
    {
      "epoch": 2.664167113015533,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 11190
    },
    {
      "epoch": 2.6644051657442125,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 11191
    },
    {
      "epoch": 2.6646432184728917,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 11192
    },
    {
      "epoch": 2.6648812712015713,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4181,
      "step": 11193
    },
    {
      "epoch": 2.6651193239302504,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4336,
      "step": 11194
    },
    {
      "epoch": 2.66535737665893,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 11195
    },
    {
      "epoch": 2.6655954293876096,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3079,
      "step": 11196
    },
    {
      "epoch": 2.665833482116289,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 11197
    },
    {
      "epoch": 2.666071534844968,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.4247,
      "step": 11198
    },
    {
      "epoch": 2.6663095875736476,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 11199
    },
    {
      "epoch": 2.666547640302327,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 11200
    },
    {
      "epoch": 2.666547640302327,
      "eval_runtime": 0.5181,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 11200
    },
    {
      "epoch": 2.6667856930310063,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 11201
    },
    {
      "epoch": 2.667023745759686,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3376,
      "step": 11202
    },
    {
      "epoch": 2.667261798488365,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 11203
    },
    {
      "epoch": 2.6674998512170447,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 11204
    },
    {
      "epoch": 2.667737903945724,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 11205
    },
    {
      "epoch": 2.6679759566744035,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4742,
      "step": 11206
    },
    {
      "epoch": 2.6682140094030826,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 11207
    },
    {
      "epoch": 2.6684520621317622,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 11208
    },
    {
      "epoch": 2.668690114860442,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 11209
    },
    {
      "epoch": 2.668928167589121,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 11210
    },
    {
      "epoch": 2.6691662203178,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 11211
    },
    {
      "epoch": 2.6694042730464798,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 11212
    },
    {
      "epoch": 2.6696423257751594,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 11213
    },
    {
      "epoch": 2.6698803785038385,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.355,
      "step": 11214
    },
    {
      "epoch": 2.670118431232518,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 11215
    },
    {
      "epoch": 2.6703564839611973,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 11216
    },
    {
      "epoch": 2.670594536689877,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5009,
      "step": 11217
    },
    {
      "epoch": 2.670832589418556,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 11218
    },
    {
      "epoch": 2.6710706421472357,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 11219
    },
    {
      "epoch": 2.671308694875915,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 11220
    },
    {
      "epoch": 2.6715467476045944,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 11221
    },
    {
      "epoch": 2.671784800333274,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 11222
    },
    {
      "epoch": 2.672022853061953,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 11223
    },
    {
      "epoch": 2.6722609057906324,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4619,
      "step": 11224
    },
    {
      "epoch": 2.672498958519312,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 11225
    },
    {
      "epoch": 2.6727370112479916,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 11226
    },
    {
      "epoch": 2.6729750639766707,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 11227
    },
    {
      "epoch": 2.6732131167053503,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 11228
    },
    {
      "epoch": 2.6734511694340295,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 11229
    },
    {
      "epoch": 2.673689222162709,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 11230
    },
    {
      "epoch": 2.6739272748913887,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 11231
    },
    {
      "epoch": 2.674165327620068,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 11232
    },
    {
      "epoch": 2.674403380348747,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 11233
    },
    {
      "epoch": 2.6746414330774266,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 11234
    },
    {
      "epoch": 2.6748794858061062,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 11235
    },
    {
      "epoch": 2.6751175385347854,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 11236
    },
    {
      "epoch": 2.675355591263465,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 11237
    },
    {
      "epoch": 2.675593643992144,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 11238
    },
    {
      "epoch": 2.6758316967208238,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 11239
    },
    {
      "epoch": 2.676069749449503,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 11240
    },
    {
      "epoch": 2.6763078021781825,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.5174,
      "step": 11241
    },
    {
      "epoch": 2.6765458549068617,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 11242
    },
    {
      "epoch": 2.6767839076355413,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 11243
    },
    {
      "epoch": 2.677021960364221,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 11244
    },
    {
      "epoch": 2.6772600130929,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.5609,
      "step": 11245
    },
    {
      "epoch": 2.6774980658215792,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 11246
    },
    {
      "epoch": 2.677736118550259,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 11247
    },
    {
      "epoch": 2.6779741712789384,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 11248
    },
    {
      "epoch": 2.6782122240076176,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3704,
      "step": 11249
    },
    {
      "epoch": 2.678450276736297,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.496,
      "step": 11250
    },
    {
      "epoch": 2.6786883294649764,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 11251
    },
    {
      "epoch": 2.678926382193656,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 11252
    },
    {
      "epoch": 2.679164434922335,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 11253
    },
    {
      "epoch": 2.6794024876510147,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 11254
    },
    {
      "epoch": 2.679640540379694,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 11255
    },
    {
      "epoch": 2.6798785931083735,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 11256
    },
    {
      "epoch": 2.680116645837053,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 11257
    },
    {
      "epoch": 2.6803546985657323,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 11258
    },
    {
      "epoch": 2.6805927512944114,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 11259
    },
    {
      "epoch": 2.680830804023091,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 11260
    },
    {
      "epoch": 2.6810688567517706,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 11261
    },
    {
      "epoch": 2.68130690948045,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4126,
      "step": 11262
    },
    {
      "epoch": 2.6815449622091294,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 11263
    },
    {
      "epoch": 2.6817830149378086,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 11264
    },
    {
      "epoch": 2.682021067666488,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 11265
    },
    {
      "epoch": 2.682259120395168,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 11266
    },
    {
      "epoch": 2.682497173123847,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 11267
    },
    {
      "epoch": 2.682735225852526,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.5001,
      "step": 11268
    },
    {
      "epoch": 2.6829732785812057,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4171,
      "step": 11269
    },
    {
      "epoch": 2.6832113313098853,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 11270
    },
    {
      "epoch": 2.6834493840385645,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 11271
    },
    {
      "epoch": 2.683687436767244,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 11272
    },
    {
      "epoch": 2.6839254894959232,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 11273
    },
    {
      "epoch": 2.684163542224603,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 11274
    },
    {
      "epoch": 2.684401594953282,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 11275
    },
    {
      "epoch": 2.6846396476819616,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 11276
    },
    {
      "epoch": 2.6848777004106408,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 11277
    },
    {
      "epoch": 2.6851157531393204,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 11278
    },
    {
      "epoch": 2.685353805868,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 11279
    },
    {
      "epoch": 2.685591858596679,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 11280
    },
    {
      "epoch": 2.6858299113253583,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 11281
    },
    {
      "epoch": 2.686067964054038,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 11282
    },
    {
      "epoch": 2.6863060167827175,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 11283
    },
    {
      "epoch": 2.6865440695113967,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 11284
    },
    {
      "epoch": 2.6867821222400763,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 11285
    },
    {
      "epoch": 2.6870201749687554,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 11286
    },
    {
      "epoch": 2.687258227697435,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 11287
    },
    {
      "epoch": 2.6874962804261147,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 11288
    },
    {
      "epoch": 2.687734333154794,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3032,
      "step": 11289
    },
    {
      "epoch": 2.687972385883473,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 11290
    },
    {
      "epoch": 2.6882104386121526,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 11291
    },
    {
      "epoch": 2.688448491340832,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 11292
    },
    {
      "epoch": 2.6886865440695114,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 11293
    },
    {
      "epoch": 2.688924596798191,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 11294
    },
    {
      "epoch": 2.68916264952687,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3796,
      "step": 11295
    },
    {
      "epoch": 2.6894007022555497,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 11296
    },
    {
      "epoch": 2.689638754984229,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 11297
    },
    {
      "epoch": 2.6898768077129085,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2219,
      "step": 11298
    },
    {
      "epoch": 2.6901148604415877,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 11299
    },
    {
      "epoch": 2.6903529131702673,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 11300
    },
    {
      "epoch": 2.6903529131702673,
      "eval_runtime": 0.5214,
      "eval_samples_per_second": 1.918,
      "eval_steps_per_second": 1.918,
      "step": 11300
    },
    {
      "epoch": 2.690590965898947,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 11301
    },
    {
      "epoch": 2.690829018627626,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 11302
    },
    {
      "epoch": 2.691067071356305,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 11303
    },
    {
      "epoch": 2.691305124084985,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.6134,
      "step": 11304
    },
    {
      "epoch": 2.6915431768136644,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 11305
    },
    {
      "epoch": 2.6917812295423436,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 11306
    },
    {
      "epoch": 2.692019282271023,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 11307
    },
    {
      "epoch": 2.6922573349997023,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 11308
    },
    {
      "epoch": 2.692495387728382,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 11309
    },
    {
      "epoch": 2.692733440457061,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 11310
    },
    {
      "epoch": 2.6929714931857407,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 11311
    },
    {
      "epoch": 2.69320954591442,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 11312
    },
    {
      "epoch": 2.6934475986430995,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 11313
    },
    {
      "epoch": 2.693685651371779,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 11314
    },
    {
      "epoch": 2.6939237041004582,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.7532,
      "step": 11315
    },
    {
      "epoch": 2.6941617568291374,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 11316
    },
    {
      "epoch": 2.694399809557817,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2428,
      "step": 11317
    },
    {
      "epoch": 2.6946378622864966,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 11318
    },
    {
      "epoch": 2.6948759150151758,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 11319
    },
    {
      "epoch": 2.6951139677438554,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 11320
    },
    {
      "epoch": 2.6953520204725345,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 11321
    },
    {
      "epoch": 2.695590073201214,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 11322
    },
    {
      "epoch": 2.6958281259298937,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2485,
      "step": 11323
    },
    {
      "epoch": 2.696066178658573,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 11324
    },
    {
      "epoch": 2.696304231387252,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8641,
      "step": 11325
    },
    {
      "epoch": 2.6965422841159317,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 11326
    },
    {
      "epoch": 2.6967803368446113,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 11327
    },
    {
      "epoch": 2.6970183895732904,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 11328
    },
    {
      "epoch": 2.69725644230197,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 11329
    },
    {
      "epoch": 2.697494495030649,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 11330
    },
    {
      "epoch": 2.697732547759329,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 11331
    },
    {
      "epoch": 2.697970600488008,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 11332
    },
    {
      "epoch": 2.6982086532166876,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4914,
      "step": 11333
    },
    {
      "epoch": 2.6984467059453667,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 11334
    },
    {
      "epoch": 2.6986847586740463,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 11335
    },
    {
      "epoch": 2.698922811402726,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 11336
    },
    {
      "epoch": 2.699160864131405,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2846,
      "step": 11337
    },
    {
      "epoch": 2.6993989168600843,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 11338
    },
    {
      "epoch": 2.699636969588764,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 11339
    },
    {
      "epoch": 2.6998750223174435,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 11340
    },
    {
      "epoch": 2.7001130750461226,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 11341
    },
    {
      "epoch": 2.7003511277748022,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3867,
      "step": 11342
    },
    {
      "epoch": 2.7005891805034814,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 11343
    },
    {
      "epoch": 2.700827233232161,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3552,
      "step": 11344
    },
    {
      "epoch": 2.70106528596084,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 11345
    },
    {
      "epoch": 2.7013033386895198,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 11346
    },
    {
      "epoch": 2.701541391418199,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 11347
    },
    {
      "epoch": 2.7017794441468785,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 11348
    },
    {
      "epoch": 2.702017496875558,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 11349
    },
    {
      "epoch": 2.7022555496042373,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 11350
    },
    {
      "epoch": 2.7024936023329165,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 11351
    },
    {
      "epoch": 2.702731655061596,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 11352
    },
    {
      "epoch": 2.7029697077902757,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 11353
    },
    {
      "epoch": 2.703207760518955,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 11354
    },
    {
      "epoch": 2.7034458132476344,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 11355
    },
    {
      "epoch": 2.7036838659763136,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 11356
    },
    {
      "epoch": 2.703921918704993,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 11357
    },
    {
      "epoch": 2.704159971433673,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 11358
    },
    {
      "epoch": 2.704398024162352,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 11359
    },
    {
      "epoch": 2.704636076891031,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 11360
    },
    {
      "epoch": 2.7048741296197107,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 11361
    },
    {
      "epoch": 2.7051121823483903,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 11362
    },
    {
      "epoch": 2.7053502350770695,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 11363
    },
    {
      "epoch": 2.705588287805749,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 11364
    },
    {
      "epoch": 2.7058263405344283,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3528,
      "step": 11365
    },
    {
      "epoch": 2.706064393263108,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 11366
    },
    {
      "epoch": 2.706302445991787,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 11367
    },
    {
      "epoch": 2.7065404987204666,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2463,
      "step": 11368
    },
    {
      "epoch": 2.706778551449146,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 11369
    },
    {
      "epoch": 2.7070166041778254,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 11370
    },
    {
      "epoch": 2.707254656906505,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 11371
    },
    {
      "epoch": 2.707492709635184,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 11372
    },
    {
      "epoch": 2.7077307623638633,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 11373
    },
    {
      "epoch": 2.707968815092543,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 11374
    },
    {
      "epoch": 2.7082068678212226,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.8135,
      "step": 11375
    },
    {
      "epoch": 2.7084449205499017,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 11376
    },
    {
      "epoch": 2.7086829732785813,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 11377
    },
    {
      "epoch": 2.7089210260072605,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 11378
    },
    {
      "epoch": 2.70915907873594,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 11379
    },
    {
      "epoch": 2.7093971314646197,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 11380
    },
    {
      "epoch": 2.709635184193299,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 11381
    },
    {
      "epoch": 2.709873236921978,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 11382
    },
    {
      "epoch": 2.7101112896506576,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4189,
      "step": 11383
    },
    {
      "epoch": 2.7103493423793372,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.417,
      "step": 11384
    },
    {
      "epoch": 2.7105873951080164,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 11385
    },
    {
      "epoch": 2.710825447836696,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 11386
    },
    {
      "epoch": 2.711063500565375,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 11387
    },
    {
      "epoch": 2.7113015532940548,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 11388
    },
    {
      "epoch": 2.711539606022734,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3371,
      "step": 11389
    },
    {
      "epoch": 2.7117776587514135,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 11390
    },
    {
      "epoch": 2.7120157114800927,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 11391
    },
    {
      "epoch": 2.7122537642087723,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 11392
    },
    {
      "epoch": 2.712491816937452,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3291,
      "step": 11393
    },
    {
      "epoch": 2.712729869666131,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 11394
    },
    {
      "epoch": 2.71296792239481,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3302,
      "step": 11395
    },
    {
      "epoch": 2.71320597512349,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2999,
      "step": 11396
    },
    {
      "epoch": 2.7134440278521694,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 11397
    },
    {
      "epoch": 2.7136820805808486,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3696,
      "step": 11398
    },
    {
      "epoch": 2.713920133309528,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4024,
      "step": 11399
    },
    {
      "epoch": 2.7141581860382074,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3025,
      "step": 11400
    },
    {
      "epoch": 2.7141581860382074,
      "eval_runtime": 0.5247,
      "eval_samples_per_second": 1.906,
      "eval_steps_per_second": 1.906,
      "step": 11400
    },
    {
      "epoch": 2.714396238766887,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 11401
    },
    {
      "epoch": 2.714634291495566,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 11402
    },
    {
      "epoch": 2.7148723442242457,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2822,
      "step": 11403
    },
    {
      "epoch": 2.715110396952925,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 11404
    },
    {
      "epoch": 2.7153484496816045,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 11405
    },
    {
      "epoch": 2.715586502410284,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 11406
    },
    {
      "epoch": 2.7158245551389633,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 11407
    },
    {
      "epoch": 2.7160626078676424,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 11408
    },
    {
      "epoch": 2.716300660596322,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 11409
    },
    {
      "epoch": 2.7165387133250016,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 11410
    },
    {
      "epoch": 2.716776766053681,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 11411
    },
    {
      "epoch": 2.7170148187823604,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 11412
    },
    {
      "epoch": 2.7172528715110396,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 11413
    },
    {
      "epoch": 2.717490924239719,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 11414
    },
    {
      "epoch": 2.7177289769683988,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3004,
      "step": 11415
    },
    {
      "epoch": 2.717967029697078,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 11416
    },
    {
      "epoch": 2.718205082425757,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 11417
    },
    {
      "epoch": 2.7184431351544367,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3403,
      "step": 11418
    },
    {
      "epoch": 2.7186811878831163,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 11419
    },
    {
      "epoch": 2.7189192406117955,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.45,
      "step": 11420
    },
    {
      "epoch": 2.719157293340475,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 11421
    },
    {
      "epoch": 2.7193953460691542,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 11422
    },
    {
      "epoch": 2.719633398797834,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 11423
    },
    {
      "epoch": 2.719871451526513,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 11424
    },
    {
      "epoch": 2.7201095042551926,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 11425
    },
    {
      "epoch": 2.7203475569838718,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3955,
      "step": 11426
    },
    {
      "epoch": 2.7205856097125514,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 11427
    },
    {
      "epoch": 2.720823662441231,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 11428
    },
    {
      "epoch": 2.72106171516991,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 11429
    },
    {
      "epoch": 2.7212997678985893,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 11430
    },
    {
      "epoch": 2.721537820627269,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 11431
    },
    {
      "epoch": 2.7217758733559485,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 11432
    },
    {
      "epoch": 2.7220139260846277,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3431,
      "step": 11433
    },
    {
      "epoch": 2.7222519788133073,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3219,
      "step": 11434
    },
    {
      "epoch": 2.7224900315419864,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3378,
      "step": 11435
    },
    {
      "epoch": 2.722728084270666,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 11436
    },
    {
      "epoch": 2.722966136999345,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 11437
    },
    {
      "epoch": 2.723204189728025,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 11438
    },
    {
      "epoch": 2.723442242456704,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3075,
      "step": 11439
    },
    {
      "epoch": 2.7236802951853836,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 11440
    },
    {
      "epoch": 2.723918347914063,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 11441
    },
    {
      "epoch": 2.7241564006427423,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 11442
    },
    {
      "epoch": 2.7243944533714215,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 11443
    },
    {
      "epoch": 2.724632506100101,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4876,
      "step": 11444
    },
    {
      "epoch": 2.7248705588287807,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 11445
    },
    {
      "epoch": 2.72510861155746,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 11446
    },
    {
      "epoch": 2.7253466642861395,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 11447
    },
    {
      "epoch": 2.7255847170148186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5286,
      "step": 11448
    },
    {
      "epoch": 2.7258227697434982,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 11449
    },
    {
      "epoch": 2.726060822472178,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 11450
    },
    {
      "epoch": 2.726298875200857,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 11451
    },
    {
      "epoch": 2.726536927929536,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 11452
    },
    {
      "epoch": 2.7267749806582158,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 11453
    },
    {
      "epoch": 2.7270130333868954,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 11454
    },
    {
      "epoch": 2.7272510861155745,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 11455
    },
    {
      "epoch": 2.727489138844254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 11456
    },
    {
      "epoch": 2.7277271915729333,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.6095,
      "step": 11457
    },
    {
      "epoch": 2.727965244301613,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 11458
    },
    {
      "epoch": 2.728203297030292,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 11459
    },
    {
      "epoch": 2.7284413497589717,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.3619,
      "step": 11460
    },
    {
      "epoch": 2.728679402487651,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 11461
    },
    {
      "epoch": 2.7289174552163304,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 11462
    },
    {
      "epoch": 2.72915550794501,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 11463
    },
    {
      "epoch": 2.729393560673689,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4203,
      "step": 11464
    },
    {
      "epoch": 2.7296316134023684,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 11465
    },
    {
      "epoch": 2.729869666131048,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 11466
    },
    {
      "epoch": 2.7301077188597276,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4062,
      "step": 11467
    },
    {
      "epoch": 2.7303457715884067,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4053,
      "step": 11468
    },
    {
      "epoch": 2.7305838243170863,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 11469
    },
    {
      "epoch": 2.7308218770457655,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 11470
    },
    {
      "epoch": 2.731059929774445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 11471
    },
    {
      "epoch": 2.7312979825031243,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 11472
    },
    {
      "epoch": 2.731536035231804,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3059,
      "step": 11473
    },
    {
      "epoch": 2.731774087960483,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 11474
    },
    {
      "epoch": 2.7320121406891626,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 11475
    },
    {
      "epoch": 2.7322501934178423,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 11476
    },
    {
      "epoch": 2.7324882461465214,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 11477
    },
    {
      "epoch": 2.732726298875201,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4058,
      "step": 11478
    },
    {
      "epoch": 2.73296435160388,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 11479
    },
    {
      "epoch": 2.73320240433256,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 11480
    },
    {
      "epoch": 2.733440457061239,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 11481
    },
    {
      "epoch": 2.7336785097899186,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 11482
    },
    {
      "epoch": 2.7339165625185977,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 11483
    },
    {
      "epoch": 2.7341546152472773,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 11484
    },
    {
      "epoch": 2.734392667975957,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3947,
      "step": 11485
    },
    {
      "epoch": 2.734630720704636,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 11486
    },
    {
      "epoch": 2.7348687734333152,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 11487
    },
    {
      "epoch": 2.735106826161995,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3787,
      "step": 11488
    },
    {
      "epoch": 2.7353448788906745,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7249,
      "step": 11489
    },
    {
      "epoch": 2.7355829316193536,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 11490
    },
    {
      "epoch": 2.7358209843480332,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 11491
    },
    {
      "epoch": 2.7360590370767124,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 11492
    },
    {
      "epoch": 2.736297089805392,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 11493
    },
    {
      "epoch": 2.736535142534071,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 11494
    },
    {
      "epoch": 2.7367731952627508,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 11495
    },
    {
      "epoch": 2.73701124799143,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 11496
    },
    {
      "epoch": 2.7372493007201095,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 11497
    },
    {
      "epoch": 2.737487353448789,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 11498
    },
    {
      "epoch": 2.7377254061774683,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4146,
      "step": 11499
    },
    {
      "epoch": 2.7379634589061475,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 11500
    },
    {
      "epoch": 2.7379634589061475,
      "eval_runtime": 0.5193,
      "eval_samples_per_second": 1.926,
      "eval_steps_per_second": 1.926,
      "step": 11500
    },
    {
      "epoch": 2.738201511634827,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 11501
    },
    {
      "epoch": 2.7384395643635067,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3924,
      "step": 11502
    },
    {
      "epoch": 2.738677617092186,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 11503
    },
    {
      "epoch": 2.7389156698208654,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 11504
    },
    {
      "epoch": 2.7391537225495446,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 11505
    },
    {
      "epoch": 2.739391775278224,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 11506
    },
    {
      "epoch": 2.739629828006904,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 11507
    },
    {
      "epoch": 2.739867880735583,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 11508
    },
    {
      "epoch": 2.740105933464262,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 11509
    },
    {
      "epoch": 2.7403439861929417,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 11510
    },
    {
      "epoch": 2.7405820389216213,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 11511
    },
    {
      "epoch": 2.7408200916503005,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 11512
    },
    {
      "epoch": 2.74105814437898,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 11513
    },
    {
      "epoch": 2.7412961971076593,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3944,
      "step": 11514
    },
    {
      "epoch": 2.741534249836339,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 11515
    },
    {
      "epoch": 2.741772302565018,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 11516
    },
    {
      "epoch": 2.7420103552936976,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 11517
    },
    {
      "epoch": 2.742248408022377,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 11518
    },
    {
      "epoch": 2.7424864607510564,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4219,
      "step": 11519
    },
    {
      "epoch": 2.742724513479736,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2603,
      "step": 11520
    },
    {
      "epoch": 2.742962566208415,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 11521
    },
    {
      "epoch": 2.7432006189370943,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 11522
    },
    {
      "epoch": 2.743438671665774,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 11523
    },
    {
      "epoch": 2.7436767243944535,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3118,
      "step": 11524
    },
    {
      "epoch": 2.7439147771231327,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 11525
    },
    {
      "epoch": 2.7441528298518123,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 11526
    },
    {
      "epoch": 2.7443908825804915,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 11527
    },
    {
      "epoch": 2.744628935309171,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 11528
    },
    {
      "epoch": 2.7448669880378502,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3208,
      "step": 11529
    },
    {
      "epoch": 2.74510504076653,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2753,
      "step": 11530
    },
    {
      "epoch": 2.745343093495209,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 11531
    },
    {
      "epoch": 2.7455811462238886,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 11532
    },
    {
      "epoch": 2.745819198952568,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 11533
    },
    {
      "epoch": 2.7460572516812474,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 11534
    },
    {
      "epoch": 2.7462953044099265,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 11535
    },
    {
      "epoch": 2.746533357138606,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8601,
      "step": 11536
    },
    {
      "epoch": 2.7467714098672857,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5484,
      "step": 11537
    },
    {
      "epoch": 2.747009462595965,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 11538
    },
    {
      "epoch": 2.7472475153246445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 11539
    },
    {
      "epoch": 2.7474855680533237,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 11540
    },
    {
      "epoch": 2.7477236207820033,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 11541
    },
    {
      "epoch": 2.747961673510683,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4153,
      "step": 11542
    },
    {
      "epoch": 2.748199726239362,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3192,
      "step": 11543
    },
    {
      "epoch": 2.748437778968041,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3279,
      "step": 11544
    },
    {
      "epoch": 2.748675831696721,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 11545
    },
    {
      "epoch": 2.7489138844254004,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 11546
    },
    {
      "epoch": 2.7491519371540796,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 11547
    },
    {
      "epoch": 2.749389989882759,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 11548
    },
    {
      "epoch": 2.7496280426114383,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 11549
    },
    {
      "epoch": 2.749866095340118,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 11550
    },
    {
      "epoch": 2.750104148068797,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 11551
    },
    {
      "epoch": 2.7503422007974767,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3142,
      "step": 11552
    },
    {
      "epoch": 2.750580253526156,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 11553
    },
    {
      "epoch": 2.7508183062548355,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4113,
      "step": 11554
    },
    {
      "epoch": 2.751056358983515,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 11555
    },
    {
      "epoch": 2.7512944117121942,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 11556
    },
    {
      "epoch": 2.7515324644408734,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 11557
    },
    {
      "epoch": 2.751770517169553,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 11558
    },
    {
      "epoch": 2.7520085698982326,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 11559
    },
    {
      "epoch": 2.7522466226269118,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 11560
    },
    {
      "epoch": 2.7524846753555914,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 11561
    },
    {
      "epoch": 2.7527227280842705,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 11562
    },
    {
      "epoch": 2.75296078081295,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 11563
    },
    {
      "epoch": 2.7531988335416293,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3806,
      "step": 11564
    },
    {
      "epoch": 2.753436886270309,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 11565
    },
    {
      "epoch": 2.753674938998988,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 11566
    },
    {
      "epoch": 2.7539129917276677,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 11567
    },
    {
      "epoch": 2.7541510444563473,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 11568
    },
    {
      "epoch": 2.7543890971850264,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 11569
    },
    {
      "epoch": 2.754627149913706,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 11570
    },
    {
      "epoch": 2.754865202642385,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 2.3815,
      "step": 11571
    },
    {
      "epoch": 2.755103255371065,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 11572
    },
    {
      "epoch": 2.755341308099744,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 11573
    },
    {
      "epoch": 2.7555793608284236,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3941,
      "step": 11574
    },
    {
      "epoch": 2.7558174135571027,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 11575
    },
    {
      "epoch": 2.7560554662857824,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.407,
      "step": 11576
    },
    {
      "epoch": 2.756293519014462,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 11577
    },
    {
      "epoch": 2.756531571743141,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 11578
    },
    {
      "epoch": 2.7567696244718203,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 11579
    },
    {
      "epoch": 2.7570076772005,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3895,
      "step": 11580
    },
    {
      "epoch": 2.7572457299291795,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 11581
    },
    {
      "epoch": 2.7574837826578587,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 11582
    },
    {
      "epoch": 2.7577218353865383,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 11583
    },
    {
      "epoch": 2.7579598881152174,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 11584
    },
    {
      "epoch": 2.758197940843897,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 11585
    },
    {
      "epoch": 2.758435993572576,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 11586
    },
    {
      "epoch": 2.758674046301256,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 11587
    },
    {
      "epoch": 2.758912099029935,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 11588
    },
    {
      "epoch": 2.7591501517586146,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 11589
    },
    {
      "epoch": 2.759388204487294,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 11590
    },
    {
      "epoch": 2.7596262572159733,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 11591
    },
    {
      "epoch": 2.7598643099446525,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3013,
      "step": 11592
    },
    {
      "epoch": 2.760102362673332,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 11593
    },
    {
      "epoch": 2.7603404154020117,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 11594
    },
    {
      "epoch": 2.760578468130691,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 11595
    },
    {
      "epoch": 2.7608165208593705,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 11596
    },
    {
      "epoch": 2.7610545735880496,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4678,
      "step": 11597
    },
    {
      "epoch": 2.7612926263167292,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 11598
    },
    {
      "epoch": 2.761530679045409,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7542,
      "step": 11599
    },
    {
      "epoch": 2.761768731774088,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 11600
    },
    {
      "epoch": 2.761768731774088,
      "eval_runtime": 0.5209,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 11600
    },
    {
      "epoch": 2.762006784502767,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 11601
    },
    {
      "epoch": 2.7622448372314468,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 11602
    },
    {
      "epoch": 2.7624828899601264,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.4738,
      "step": 11603
    },
    {
      "epoch": 2.7627209426888055,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 11604
    },
    {
      "epoch": 2.762958995417485,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 11605
    },
    {
      "epoch": 2.7631970481461643,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 11606
    },
    {
      "epoch": 2.763435100874844,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 11607
    },
    {
      "epoch": 2.763673153603523,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 11608
    },
    {
      "epoch": 2.7639112063322027,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 11609
    },
    {
      "epoch": 2.764149259060882,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 11610
    },
    {
      "epoch": 2.7643873117895614,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 11611
    },
    {
      "epoch": 2.764625364518241,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 11612
    },
    {
      "epoch": 2.76486341724692,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 11613
    },
    {
      "epoch": 2.7651014699755994,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 11614
    },
    {
      "epoch": 2.765339522704279,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 11615
    },
    {
      "epoch": 2.7655775754329586,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 11616
    },
    {
      "epoch": 2.7658156281616377,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3231,
      "step": 11617
    },
    {
      "epoch": 2.7660536808903173,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3055,
      "step": 11618
    },
    {
      "epoch": 2.7662917336189965,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 11619
    },
    {
      "epoch": 2.766529786347676,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 11620
    },
    {
      "epoch": 2.7667678390763553,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 11621
    },
    {
      "epoch": 2.767005891805035,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 11622
    },
    {
      "epoch": 2.767243944533714,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 11623
    },
    {
      "epoch": 2.7674819972623936,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 11624
    },
    {
      "epoch": 2.7677200499910732,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 11625
    },
    {
      "epoch": 2.7679581027197524,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 11626
    },
    {
      "epoch": 2.7681961554484316,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 11627
    },
    {
      "epoch": 2.768434208177111,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 11628
    },
    {
      "epoch": 2.7686722609057908,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 11629
    },
    {
      "epoch": 2.76891031363447,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3724,
      "step": 11630
    },
    {
      "epoch": 2.7691483663631495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 11631
    },
    {
      "epoch": 2.7693864190918287,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 11632
    },
    {
      "epoch": 2.7696244718205083,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 11633
    },
    {
      "epoch": 2.769862524549188,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 11634
    },
    {
      "epoch": 2.770100577277867,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 11635
    },
    {
      "epoch": 2.7703386300065462,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 11636
    },
    {
      "epoch": 2.770576682735226,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 11637
    },
    {
      "epoch": 2.7708147354639054,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 11638
    },
    {
      "epoch": 2.7710527881925846,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 11639
    },
    {
      "epoch": 2.771290840921264,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 11640
    },
    {
      "epoch": 2.7715288936499434,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 11641
    },
    {
      "epoch": 2.771766946378623,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3956,
      "step": 11642
    },
    {
      "epoch": 2.772004999107302,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 11643
    },
    {
      "epoch": 2.7722430518359817,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 11644
    },
    {
      "epoch": 2.772481104564661,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.4202,
      "step": 11645
    },
    {
      "epoch": 2.7727191572933405,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 11646
    },
    {
      "epoch": 2.77295721002202,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 11647
    },
    {
      "epoch": 2.7731952627506993,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 11648
    },
    {
      "epoch": 2.7734333154793784,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 11649
    },
    {
      "epoch": 2.773671368208058,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 11650
    },
    {
      "epoch": 2.7739094209367376,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 11651
    },
    {
      "epoch": 2.774147473665417,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 11652
    },
    {
      "epoch": 2.7743855263940964,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 11653
    },
    {
      "epoch": 2.7746235791227756,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4972,
      "step": 11654
    },
    {
      "epoch": 2.774861631851455,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 11655
    },
    {
      "epoch": 2.7750996845801343,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 11656
    },
    {
      "epoch": 2.775337737308814,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3939,
      "step": 11657
    },
    {
      "epoch": 2.775575790037493,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 11658
    },
    {
      "epoch": 2.7758138427661727,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 11659
    },
    {
      "epoch": 2.7760518954948523,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 11660
    },
    {
      "epoch": 2.7762899482235315,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4874,
      "step": 11661
    },
    {
      "epoch": 2.7765280009522106,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 11662
    },
    {
      "epoch": 2.7767660536808902,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 11663
    },
    {
      "epoch": 2.77700410640957,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 11664
    },
    {
      "epoch": 2.777242159138249,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 11665
    },
    {
      "epoch": 2.7774802118669286,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 11666
    },
    {
      "epoch": 2.7777182645956078,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 11667
    },
    {
      "epoch": 2.7779563173242874,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 11668
    },
    {
      "epoch": 2.778194370052967,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 11669
    },
    {
      "epoch": 2.778432422781646,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 11670
    },
    {
      "epoch": 2.7786704755103253,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 11671
    },
    {
      "epoch": 2.778908528239005,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 11672
    },
    {
      "epoch": 2.7791465809676845,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 11673
    },
    {
      "epoch": 2.7793846336963637,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 11674
    },
    {
      "epoch": 2.7796226864250433,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5298,
      "step": 11675
    },
    {
      "epoch": 2.7798607391537224,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 11676
    },
    {
      "epoch": 2.780098791882402,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 11677
    },
    {
      "epoch": 2.780336844611081,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 11678
    },
    {
      "epoch": 2.780574897339761,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.363,
      "step": 11679
    },
    {
      "epoch": 2.78081295006844,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 11680
    },
    {
      "epoch": 2.7810510027971196,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 11681
    },
    {
      "epoch": 2.781289055525799,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 11682
    },
    {
      "epoch": 2.7815271082544784,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 11683
    },
    {
      "epoch": 2.7817651609831575,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2969,
      "step": 11684
    },
    {
      "epoch": 2.782003213711837,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 11685
    },
    {
      "epoch": 2.7822412664405167,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2788,
      "step": 11686
    },
    {
      "epoch": 2.782479319169196,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 11687
    },
    {
      "epoch": 2.7827173718978755,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 11688
    },
    {
      "epoch": 2.7829554246265547,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2663,
      "step": 11689
    },
    {
      "epoch": 2.7831934773552343,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 11690
    },
    {
      "epoch": 2.783431530083914,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 11691
    },
    {
      "epoch": 2.783669582812593,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 11692
    },
    {
      "epoch": 2.783907635541272,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 11693
    },
    {
      "epoch": 2.784145688269952,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 11694
    },
    {
      "epoch": 2.7843837409986314,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 11695
    },
    {
      "epoch": 2.7846217937273106,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 11696
    },
    {
      "epoch": 2.78485984645599,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 11697
    },
    {
      "epoch": 2.7850978991846693,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 11698
    },
    {
      "epoch": 2.785335951913349,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0575,
      "step": 11699
    },
    {
      "epoch": 2.785574004642028,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 11700
    },
    {
      "epoch": 2.785574004642028,
      "eval_runtime": 0.5293,
      "eval_samples_per_second": 1.889,
      "eval_steps_per_second": 1.889,
      "step": 11700
    },
    {
      "epoch": 2.7858120573707077,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 11701
    },
    {
      "epoch": 2.786050110099387,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.4906,
      "step": 11702
    },
    {
      "epoch": 2.7862881628280665,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 11703
    },
    {
      "epoch": 2.786526215556746,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 11704
    },
    {
      "epoch": 2.7867642682854252,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 11705
    },
    {
      "epoch": 2.7870023210141044,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 11706
    },
    {
      "epoch": 2.787240373742784,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 11707
    },
    {
      "epoch": 2.7874784264714636,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 11708
    },
    {
      "epoch": 2.7877164792001428,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 11709
    },
    {
      "epoch": 2.7879545319288224,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5557,
      "step": 11710
    },
    {
      "epoch": 2.7881925846575015,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4698,
      "step": 11711
    },
    {
      "epoch": 2.788430637386181,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 11712
    },
    {
      "epoch": 2.7886686901148603,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3525,
      "step": 11713
    },
    {
      "epoch": 2.78890674284354,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 11714
    },
    {
      "epoch": 2.789144795572219,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3414,
      "step": 11715
    },
    {
      "epoch": 2.7893828483008987,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 11716
    },
    {
      "epoch": 2.7896209010295783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4943,
      "step": 11717
    },
    {
      "epoch": 2.7898589537582574,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 11718
    },
    {
      "epoch": 2.7900970064869366,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 11719
    },
    {
      "epoch": 2.790335059215616,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 11720
    },
    {
      "epoch": 2.790573111944296,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 11721
    },
    {
      "epoch": 2.790811164672975,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 11722
    },
    {
      "epoch": 2.7910492174016546,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 11723
    },
    {
      "epoch": 2.7912872701303337,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3906,
      "step": 11724
    },
    {
      "epoch": 2.7915253228590133,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 11725
    },
    {
      "epoch": 2.791763375587693,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 11726
    },
    {
      "epoch": 2.792001428316372,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3373,
      "step": 11727
    },
    {
      "epoch": 2.7922394810450513,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 11728
    },
    {
      "epoch": 2.792477533773731,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 11729
    },
    {
      "epoch": 2.7927155865024105,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 11730
    },
    {
      "epoch": 2.7929536392310896,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 11731
    },
    {
      "epoch": 2.7931916919597692,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 11732
    },
    {
      "epoch": 2.7934297446884484,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 11733
    },
    {
      "epoch": 2.793667797417128,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 11734
    },
    {
      "epoch": 2.793905850145807,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4461,
      "step": 11735
    },
    {
      "epoch": 2.7941439028744868,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 11736
    },
    {
      "epoch": 2.794381955603166,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 11737
    },
    {
      "epoch": 2.7946200083318455,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 11738
    },
    {
      "epoch": 2.794858061060525,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.316,
      "step": 11739
    },
    {
      "epoch": 2.7950961137892043,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 11740
    },
    {
      "epoch": 2.7953341665178835,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 11741
    },
    {
      "epoch": 2.795572219246563,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 11742
    },
    {
      "epoch": 2.7958102719752427,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 11743
    },
    {
      "epoch": 2.796048324703922,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 11744
    },
    {
      "epoch": 2.7962863774326014,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 11745
    },
    {
      "epoch": 2.7965244301612806,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3407,
      "step": 11746
    },
    {
      "epoch": 2.79676248288996,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 11747
    },
    {
      "epoch": 2.7970005356186394,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 11748
    },
    {
      "epoch": 2.797238588347319,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 11749
    },
    {
      "epoch": 2.797476641075998,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 11750
    },
    {
      "epoch": 2.7977146938046777,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 11751
    },
    {
      "epoch": 2.7979527465333573,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 11752
    },
    {
      "epoch": 2.7981907992620365,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 11753
    },
    {
      "epoch": 2.7984288519907157,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 11754
    },
    {
      "epoch": 2.7986669047193953,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 11755
    },
    {
      "epoch": 2.798904957448075,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3521,
      "step": 11756
    },
    {
      "epoch": 2.799143010176754,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 11757
    },
    {
      "epoch": 2.7993810629054336,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 11758
    },
    {
      "epoch": 2.799619115634113,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 11759
    },
    {
      "epoch": 2.7998571683627924,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 11760
    },
    {
      "epoch": 2.800095221091472,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 11761
    },
    {
      "epoch": 2.800333273820151,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 11762
    },
    {
      "epoch": 2.8005713265488303,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 11763
    },
    {
      "epoch": 2.80080937927751,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 11764
    },
    {
      "epoch": 2.8010474320061896,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3813,
      "step": 11765
    },
    {
      "epoch": 2.8012854847348687,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 11766
    },
    {
      "epoch": 2.8015235374635483,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3992,
      "step": 11767
    },
    {
      "epoch": 2.8017615901922275,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3089,
      "step": 11768
    },
    {
      "epoch": 2.801999642920907,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 11769
    },
    {
      "epoch": 2.8022376956495862,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 11770
    },
    {
      "epoch": 2.802475748378266,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 11771
    },
    {
      "epoch": 2.802713801106945,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 11772
    },
    {
      "epoch": 2.8029518538356246,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 11773
    },
    {
      "epoch": 2.8031899065643042,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 11774
    },
    {
      "epoch": 2.8034279592929834,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3495,
      "step": 11775
    },
    {
      "epoch": 2.8036660120216625,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 11776
    },
    {
      "epoch": 2.803904064750342,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 11777
    },
    {
      "epoch": 2.8041421174790218,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 11778
    },
    {
      "epoch": 2.804380170207701,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 11779
    },
    {
      "epoch": 2.8046182229363805,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 11780
    },
    {
      "epoch": 2.8048562756650597,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 11781
    },
    {
      "epoch": 2.8050943283937393,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 11782
    },
    {
      "epoch": 2.8053323811224185,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 11783
    },
    {
      "epoch": 2.805570433851098,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3156,
      "step": 11784
    },
    {
      "epoch": 2.805808486579777,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 11785
    },
    {
      "epoch": 2.806046539308457,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2932,
      "step": 11786
    },
    {
      "epoch": 2.8062845920371364,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 11787
    },
    {
      "epoch": 2.8065226447658156,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 11788
    },
    {
      "epoch": 2.806760697494495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 11789
    },
    {
      "epoch": 2.8069987502231744,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 11790
    },
    {
      "epoch": 2.807236802951854,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3231,
      "step": 11791
    },
    {
      "epoch": 2.807474855680533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4007,
      "step": 11792
    },
    {
      "epoch": 2.8077129084092127,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 11793
    },
    {
      "epoch": 2.807950961137892,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 11794
    },
    {
      "epoch": 2.8081890138665715,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 11795
    },
    {
      "epoch": 2.808427066595251,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 11796
    },
    {
      "epoch": 2.8086651193239303,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 11797
    },
    {
      "epoch": 2.8089031720526094,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.569,
      "step": 11798
    },
    {
      "epoch": 2.809141224781289,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 11799
    },
    {
      "epoch": 2.8093792775099686,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 11800
    },
    {
      "epoch": 2.8093792775099686,
      "eval_runtime": 0.5211,
      "eval_samples_per_second": 1.919,
      "eval_steps_per_second": 1.919,
      "step": 11800
    },
    {
      "epoch": 2.809617330238648,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 11801
    },
    {
      "epoch": 2.8098553829673274,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 11802
    },
    {
      "epoch": 2.8100934356960066,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 11803
    },
    {
      "epoch": 2.810331488424686,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2997,
      "step": 11804
    },
    {
      "epoch": 2.8105695411533653,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 11805
    },
    {
      "epoch": 2.810807593882045,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 11806
    },
    {
      "epoch": 2.811045646610724,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 11807
    },
    {
      "epoch": 2.8112836993394037,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 11808
    },
    {
      "epoch": 2.8115217520680833,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 11809
    },
    {
      "epoch": 2.8117598047967625,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.4322,
      "step": 11810
    },
    {
      "epoch": 2.8119978575254416,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 11811
    },
    {
      "epoch": 2.8122359102541212,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 11812
    },
    {
      "epoch": 2.812473962982801,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 11813
    },
    {
      "epoch": 2.81271201571148,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4883,
      "step": 11814
    },
    {
      "epoch": 2.8129500684401596,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3299,
      "step": 11815
    },
    {
      "epoch": 2.8131881211688388,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 11816
    },
    {
      "epoch": 2.8134261738975184,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 11817
    },
    {
      "epoch": 2.813664226626198,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 11818
    },
    {
      "epoch": 2.813902279354877,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9303,
      "step": 11819
    },
    {
      "epoch": 2.8141403320835563,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 11820
    },
    {
      "epoch": 2.814378384812236,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4278,
      "step": 11821
    },
    {
      "epoch": 2.8146164375409155,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.393,
      "step": 11822
    },
    {
      "epoch": 2.8148544902695947,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 11823
    },
    {
      "epoch": 2.8150925429982743,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 11824
    },
    {
      "epoch": 2.8153305957269534,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 11825
    },
    {
      "epoch": 2.815568648455633,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 11826
    },
    {
      "epoch": 2.815806701184312,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 11827
    },
    {
      "epoch": 2.816044753912992,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 11828
    },
    {
      "epoch": 2.816282806641671,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 11829
    },
    {
      "epoch": 2.8165208593703506,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 11830
    },
    {
      "epoch": 2.81675891209903,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 11831
    },
    {
      "epoch": 2.8169969648277093,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 11832
    },
    {
      "epoch": 2.8172350175563885,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 11833
    },
    {
      "epoch": 2.817473070285068,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 11834
    },
    {
      "epoch": 2.8177111230137477,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 11835
    },
    {
      "epoch": 2.817949175742427,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 11836
    },
    {
      "epoch": 2.8181872284711065,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 11837
    },
    {
      "epoch": 2.8184252811997856,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 11838
    },
    {
      "epoch": 2.8186633339284652,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 11839
    },
    {
      "epoch": 2.8189013866571444,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 11840
    },
    {
      "epoch": 2.819139439385824,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 11841
    },
    {
      "epoch": 2.819377492114503,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 11842
    },
    {
      "epoch": 2.8196155448431828,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3913,
      "step": 11843
    },
    {
      "epoch": 2.8198535975718624,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 11844
    },
    {
      "epoch": 2.8200916503005415,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 11845
    },
    {
      "epoch": 2.8203297030292207,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 11846
    },
    {
      "epoch": 2.8205677557579003,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.7889,
      "step": 11847
    },
    {
      "epoch": 2.82080580848658,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 11848
    },
    {
      "epoch": 2.821043861215259,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 11849
    },
    {
      "epoch": 2.8212819139439387,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 11850
    },
    {
      "epoch": 2.821519966672618,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 11851
    },
    {
      "epoch": 2.8217580194012974,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 11852
    },
    {
      "epoch": 2.821996072129977,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3014,
      "step": 11853
    },
    {
      "epoch": 2.822234124858656,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3006,
      "step": 11854
    },
    {
      "epoch": 2.8224721775873354,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 11855
    },
    {
      "epoch": 2.822710230316015,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 11856
    },
    {
      "epoch": 2.8229482830446946,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 11857
    },
    {
      "epoch": 2.8231863357733737,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 11858
    },
    {
      "epoch": 2.8234243885020534,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 11859
    },
    {
      "epoch": 2.8236624412307325,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 11860
    },
    {
      "epoch": 2.823900493959412,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 11861
    },
    {
      "epoch": 2.8241385466880913,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 11862
    },
    {
      "epoch": 2.824376599416771,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 11863
    },
    {
      "epoch": 2.82461465214545,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 11864
    },
    {
      "epoch": 2.8248527048741296,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 11865
    },
    {
      "epoch": 2.8250907576028093,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 11866
    },
    {
      "epoch": 2.8253288103314884,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 11867
    },
    {
      "epoch": 2.8255668630601676,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 11868
    },
    {
      "epoch": 2.825804915788847,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 11869
    },
    {
      "epoch": 2.826042968517527,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3197,
      "step": 11870
    },
    {
      "epoch": 2.826281021246206,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 11871
    },
    {
      "epoch": 2.8265190739748856,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 11872
    },
    {
      "epoch": 2.8267571267035647,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 11873
    },
    {
      "epoch": 2.8269951794322443,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 11874
    },
    {
      "epoch": 2.8272332321609235,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 11875
    },
    {
      "epoch": 2.827471284889603,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 11876
    },
    {
      "epoch": 2.8277093376182822,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 11877
    },
    {
      "epoch": 2.827947390346962,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 11878
    },
    {
      "epoch": 2.8281854430756415,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 11879
    },
    {
      "epoch": 2.8284234958043206,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3748,
      "step": 11880
    },
    {
      "epoch": 2.8286615485330002,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2828,
      "step": 11881
    },
    {
      "epoch": 2.8288996012616794,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8269,
      "step": 11882
    },
    {
      "epoch": 2.829137653990359,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 11883
    },
    {
      "epoch": 2.829375706719038,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 11884
    },
    {
      "epoch": 2.8296137594477178,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3961,
      "step": 11885
    },
    {
      "epoch": 2.829851812176397,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 11886
    },
    {
      "epoch": 2.8300898649050765,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 11887
    },
    {
      "epoch": 2.830327917633756,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3425,
      "step": 11888
    },
    {
      "epoch": 2.8305659703624353,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 11889
    },
    {
      "epoch": 2.8308040230911145,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3545,
      "step": 11890
    },
    {
      "epoch": 2.831042075819794,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 11891
    },
    {
      "epoch": 2.8312801285484737,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 11892
    },
    {
      "epoch": 2.831518181277153,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 11893
    },
    {
      "epoch": 2.8317562340058324,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 11894
    },
    {
      "epoch": 2.8319942867345116,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 11895
    },
    {
      "epoch": 2.832232339463191,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 11896
    },
    {
      "epoch": 2.8324703921918704,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 11897
    },
    {
      "epoch": 2.83270844492055,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 11898
    },
    {
      "epoch": 2.832946497649229,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3073,
      "step": 11899
    },
    {
      "epoch": 2.8331845503779087,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 11900
    },
    {
      "epoch": 2.8331845503779087,
      "eval_runtime": 0.529,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 11900
    },
    {
      "epoch": 2.8334226031065883,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 11901
    },
    {
      "epoch": 2.8336606558352675,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 11902
    },
    {
      "epoch": 2.8338987085639467,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 11903
    },
    {
      "epoch": 2.8341367612926263,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4294,
      "step": 11904
    },
    {
      "epoch": 2.834374814021306,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 11905
    },
    {
      "epoch": 2.834612866749985,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 11906
    },
    {
      "epoch": 2.8348509194786646,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 11907
    },
    {
      "epoch": 2.835088972207344,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1506,
      "step": 11908
    },
    {
      "epoch": 2.8353270249360234,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 11909
    },
    {
      "epoch": 2.835565077664703,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 11910
    },
    {
      "epoch": 2.835803130393382,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 11911
    },
    {
      "epoch": 2.8360411831220613,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 11912
    },
    {
      "epoch": 2.836279235850741,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 11913
    },
    {
      "epoch": 2.8365172885794205,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3937,
      "step": 11914
    },
    {
      "epoch": 2.8367553413080997,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 11915
    },
    {
      "epoch": 2.8369933940367793,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 11916
    },
    {
      "epoch": 2.8372314467654585,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 11917
    },
    {
      "epoch": 2.837469499494138,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 11918
    },
    {
      "epoch": 2.8377075522228172,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 11919
    },
    {
      "epoch": 2.837945604951497,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 11920
    },
    {
      "epoch": 2.838183657680176,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 11921
    },
    {
      "epoch": 2.8384217104088556,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 11922
    },
    {
      "epoch": 2.838659763137535,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4198,
      "step": 11923
    },
    {
      "epoch": 2.8388978158662144,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 11924
    },
    {
      "epoch": 2.8391358685948935,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 11925
    },
    {
      "epoch": 2.839373921323573,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.3963,
      "step": 11926
    },
    {
      "epoch": 2.8396119740522527,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 11927
    },
    {
      "epoch": 2.839850026780932,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 11928
    },
    {
      "epoch": 2.8400880795096115,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 11929
    },
    {
      "epoch": 2.8403261322382907,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 11930
    },
    {
      "epoch": 2.8405641849669703,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 11931
    },
    {
      "epoch": 2.8408022376956494,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 11932
    },
    {
      "epoch": 2.841040290424329,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5537,
      "step": 11933
    },
    {
      "epoch": 2.841278343153008,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 11934
    },
    {
      "epoch": 2.841516395881688,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 11935
    },
    {
      "epoch": 2.8417544486103674,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 11936
    },
    {
      "epoch": 2.8419925013390466,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 11937
    },
    {
      "epoch": 2.8422305540677257,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 11938
    },
    {
      "epoch": 2.8424686067964053,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1914,
      "step": 11939
    },
    {
      "epoch": 2.842706659525085,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 11940
    },
    {
      "epoch": 2.842944712253764,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 11941
    },
    {
      "epoch": 2.8431827649824437,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 11942
    },
    {
      "epoch": 2.843420817711123,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 11943
    },
    {
      "epoch": 2.8436588704398025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 11944
    },
    {
      "epoch": 2.843896923168482,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 11945
    },
    {
      "epoch": 2.8441349758971612,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 11946
    },
    {
      "epoch": 2.8443730286258404,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.7392,
      "step": 11947
    },
    {
      "epoch": 2.84461108135452,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 11948
    },
    {
      "epoch": 2.8448491340831996,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 11949
    },
    {
      "epoch": 2.8450871868118788,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 11950
    },
    {
      "epoch": 2.8453252395405584,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 11951
    },
    {
      "epoch": 2.8455632922692375,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 11952
    },
    {
      "epoch": 2.845801344997917,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 11953
    },
    {
      "epoch": 2.8460393977265963,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 11954
    },
    {
      "epoch": 2.846277450455276,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 11955
    },
    {
      "epoch": 2.846515503183955,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 11956
    },
    {
      "epoch": 2.8467535559126347,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 11957
    },
    {
      "epoch": 2.8469916086413143,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2834,
      "step": 11958
    },
    {
      "epoch": 2.8472296613699934,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 11959
    },
    {
      "epoch": 2.8474677140986726,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 11960
    },
    {
      "epoch": 2.847705766827352,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 11961
    },
    {
      "epoch": 2.847943819556032,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 11962
    },
    {
      "epoch": 2.848181872284711,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 11963
    },
    {
      "epoch": 2.8484199250133906,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 11964
    },
    {
      "epoch": 2.8486579777420697,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 11965
    },
    {
      "epoch": 2.8488960304707494,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 11966
    },
    {
      "epoch": 2.8491340831994285,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 11967
    },
    {
      "epoch": 2.849372135928108,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 11968
    },
    {
      "epoch": 2.8496101886567873,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 11969
    },
    {
      "epoch": 2.849848241385467,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 11970
    },
    {
      "epoch": 2.8500862941141465,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3767,
      "step": 11971
    },
    {
      "epoch": 2.8503243468428257,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 11972
    },
    {
      "epoch": 2.850562399571505,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2673,
      "step": 11973
    },
    {
      "epoch": 2.8508004523001844,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 11974
    },
    {
      "epoch": 2.851038505028864,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 11975
    },
    {
      "epoch": 2.851276557757543,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 11976
    },
    {
      "epoch": 2.851514610486223,
      "grad_norm": 2.546875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 11977
    },
    {
      "epoch": 2.851752663214902,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 11978
    },
    {
      "epoch": 2.8519907159435816,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 11979
    },
    {
      "epoch": 2.852228768672261,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 11980
    },
    {
      "epoch": 2.8524668214009403,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 11981
    },
    {
      "epoch": 2.8527048741296195,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4386,
      "step": 11982
    },
    {
      "epoch": 2.852942926858299,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 11983
    },
    {
      "epoch": 2.8531809795869787,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.5266,
      "step": 11984
    },
    {
      "epoch": 2.853419032315658,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 11985
    },
    {
      "epoch": 2.8536570850443375,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 11986
    },
    {
      "epoch": 2.8538951377730166,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 11987
    },
    {
      "epoch": 2.8541331905016962,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 11988
    },
    {
      "epoch": 2.8543712432303754,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 11989
    },
    {
      "epoch": 2.854609295959055,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 11990
    },
    {
      "epoch": 2.854847348687734,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 11991
    },
    {
      "epoch": 2.8550854014164138,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 11992
    },
    {
      "epoch": 2.8553234541450934,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 11993
    },
    {
      "epoch": 2.8555615068737725,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 11994
    },
    {
      "epoch": 2.8557995596024517,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 11995
    },
    {
      "epoch": 2.8560376123311313,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 11996
    },
    {
      "epoch": 2.856275665059811,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 11997
    },
    {
      "epoch": 2.85651371778849,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 11998
    },
    {
      "epoch": 2.8567517705171697,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 11999
    },
    {
      "epoch": 2.856989823245849,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 12000
    },
    {
      "epoch": 2.856989823245849,
      "eval_runtime": 0.53,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 12000
    },
    {
      "epoch": 2.8572278759745284,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 12001
    },
    {
      "epoch": 2.857465928703208,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 12002
    },
    {
      "epoch": 2.857703981431887,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3751,
      "step": 12003
    },
    {
      "epoch": 2.8579420341605664,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 12004
    },
    {
      "epoch": 2.858180086889246,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 12005
    },
    {
      "epoch": 2.8584181396179256,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 12006
    },
    {
      "epoch": 2.8586561923466047,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 12007
    },
    {
      "epoch": 2.8588942450752843,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.45,
      "step": 12008
    },
    {
      "epoch": 2.8591322978039635,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 12009
    },
    {
      "epoch": 2.859370350532643,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 12010
    },
    {
      "epoch": 2.8596084032613223,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 12011
    },
    {
      "epoch": 2.859846455990002,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 12012
    },
    {
      "epoch": 2.860084508718681,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 12013
    },
    {
      "epoch": 2.8603225614473606,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 12014
    },
    {
      "epoch": 2.8605606141760402,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 12015
    },
    {
      "epoch": 2.8607986669047194,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 12016
    },
    {
      "epoch": 2.8610367196333986,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 12017
    },
    {
      "epoch": 2.861274772362078,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 12018
    },
    {
      "epoch": 2.8615128250907578,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 12019
    },
    {
      "epoch": 2.861750877819437,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 12020
    },
    {
      "epoch": 2.8619889305481165,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 12021
    },
    {
      "epoch": 2.8622269832767957,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 12022
    },
    {
      "epoch": 2.8624650360054753,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 12023
    },
    {
      "epoch": 2.8627030887341545,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 12024
    },
    {
      "epoch": 2.862941141462834,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 12025
    },
    {
      "epoch": 2.8631791941915132,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4048,
      "step": 12026
    },
    {
      "epoch": 2.863417246920193,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 12027
    },
    {
      "epoch": 2.8636552996488724,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 12028
    },
    {
      "epoch": 2.8638933523775516,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 12029
    },
    {
      "epoch": 2.8641314051062308,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 12030
    },
    {
      "epoch": 2.8643694578349104,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 12031
    },
    {
      "epoch": 2.86460751056359,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 12032
    },
    {
      "epoch": 2.864845563292269,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 12033
    },
    {
      "epoch": 2.8650836160209487,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 12034
    },
    {
      "epoch": 2.865321668749628,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 12035
    },
    {
      "epoch": 2.8655597214783075,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 12036
    },
    {
      "epoch": 2.865797774206987,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 12037
    },
    {
      "epoch": 2.8660358269356663,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8229,
      "step": 12038
    },
    {
      "epoch": 2.8662738796643454,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 12039
    },
    {
      "epoch": 2.866511932393025,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 12040
    },
    {
      "epoch": 2.8667499851217046,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 12041
    },
    {
      "epoch": 2.866988037850384,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 12042
    },
    {
      "epoch": 2.8672260905790634,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 12043
    },
    {
      "epoch": 2.8674641433077426,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 12044
    },
    {
      "epoch": 2.867702196036422,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 12045
    },
    {
      "epoch": 2.8679402487651013,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 12046
    },
    {
      "epoch": 2.868178301493781,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 12047
    },
    {
      "epoch": 2.86841635422246,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 12048
    },
    {
      "epoch": 2.8686544069511397,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 12049
    },
    {
      "epoch": 2.8688924596798193,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 12050
    },
    {
      "epoch": 2.8691305124084985,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 12051
    },
    {
      "epoch": 2.8693685651371776,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 12052
    },
    {
      "epoch": 2.8696066178658572,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 12053
    },
    {
      "epoch": 2.869844670594537,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 12054
    },
    {
      "epoch": 2.870082723323216,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 12055
    },
    {
      "epoch": 2.8703207760518956,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3927,
      "step": 12056
    },
    {
      "epoch": 2.870558828780575,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 12057
    },
    {
      "epoch": 2.8707968815092544,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3423,
      "step": 12058
    },
    {
      "epoch": 2.8710349342379335,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 12059
    },
    {
      "epoch": 2.871272986966613,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3133,
      "step": 12060
    },
    {
      "epoch": 2.8715110396952923,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 12061
    },
    {
      "epoch": 2.871749092423972,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 12062
    },
    {
      "epoch": 2.8719871451526515,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 12063
    },
    {
      "epoch": 2.8722251978813307,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 12064
    },
    {
      "epoch": 2.87246325061001,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 12065
    },
    {
      "epoch": 2.8727013033386894,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 12066
    },
    {
      "epoch": 2.872939356067369,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 12067
    },
    {
      "epoch": 2.873177408796048,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 12068
    },
    {
      "epoch": 2.873415461524728,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 12069
    },
    {
      "epoch": 2.873653514253407,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4395,
      "step": 12070
    },
    {
      "epoch": 2.8738915669820866,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 12071
    },
    {
      "epoch": 2.874129619710766,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 12072
    },
    {
      "epoch": 2.8743676724394454,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 12073
    },
    {
      "epoch": 2.8746057251681245,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 12074
    },
    {
      "epoch": 2.874843777896804,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2269,
      "step": 12075
    },
    {
      "epoch": 2.8750818306254837,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 12076
    },
    {
      "epoch": 2.875319883354163,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.5258,
      "step": 12077
    },
    {
      "epoch": 2.8755579360828425,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 12078
    },
    {
      "epoch": 2.8757959888115217,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 12079
    },
    {
      "epoch": 2.8760340415402013,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 12080
    },
    {
      "epoch": 2.8762720942688804,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 12081
    },
    {
      "epoch": 2.87651014699756,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 12082
    },
    {
      "epoch": 2.876748199726239,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 12083
    },
    {
      "epoch": 2.876986252454919,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 12084
    },
    {
      "epoch": 2.8772243051835984,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 12085
    },
    {
      "epoch": 2.8774623579122776,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 12086
    },
    {
      "epoch": 2.8777004106409567,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 12087
    },
    {
      "epoch": 2.8779384633696363,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 12088
    },
    {
      "epoch": 2.878176516098316,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2471,
      "step": 12089
    },
    {
      "epoch": 2.878414568826995,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 12090
    },
    {
      "epoch": 2.8786526215556747,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9538,
      "step": 12091
    },
    {
      "epoch": 2.878890674284354,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 12092
    },
    {
      "epoch": 2.8791287270130335,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 12093
    },
    {
      "epoch": 2.8793667797417126,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 12094
    },
    {
      "epoch": 2.8796048324703922,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 12095
    },
    {
      "epoch": 2.8798428851990714,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 12096
    },
    {
      "epoch": 2.880080937927751,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 12097
    },
    {
      "epoch": 2.8803189906564306,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 12098
    },
    {
      "epoch": 2.8805570433851098,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 12099
    },
    {
      "epoch": 2.8807950961137894,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 12100
    },
    {
      "epoch": 2.8807950961137894,
      "eval_runtime": 0.518,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 12100
    },
    {
      "epoch": 2.8810331488424685,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 12101
    },
    {
      "epoch": 2.881271201571148,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 12102
    },
    {
      "epoch": 2.8815092542998273,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 12103
    },
    {
      "epoch": 2.881747307028507,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 12104
    },
    {
      "epoch": 2.881985359757186,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 12105
    },
    {
      "epoch": 2.8822234124858657,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 12106
    },
    {
      "epoch": 2.8824614652145453,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 12107
    },
    {
      "epoch": 2.8826995179432244,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 12108
    },
    {
      "epoch": 2.8829375706719036,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4818,
      "step": 12109
    },
    {
      "epoch": 2.883175623400583,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 12110
    },
    {
      "epoch": 2.883413676129263,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 12111
    },
    {
      "epoch": 2.883651728857942,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 12112
    },
    {
      "epoch": 2.8838897815866216,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 12113
    },
    {
      "epoch": 2.8841278343153007,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 12114
    },
    {
      "epoch": 2.8843658870439803,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 12115
    },
    {
      "epoch": 2.8846039397726595,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2604,
      "step": 12116
    },
    {
      "epoch": 2.884841992501339,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 12117
    },
    {
      "epoch": 2.8850800452300183,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 12118
    },
    {
      "epoch": 2.885318097958698,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 12119
    },
    {
      "epoch": 2.8855561506873775,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 12120
    },
    {
      "epoch": 2.8857942034160566,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 12121
    },
    {
      "epoch": 2.886032256144736,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 12122
    },
    {
      "epoch": 2.8862703088734154,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 12123
    },
    {
      "epoch": 2.886508361602095,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 12124
    },
    {
      "epoch": 2.886746414330774,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 12125
    },
    {
      "epoch": 2.8869844670594538,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.414,
      "step": 12126
    },
    {
      "epoch": 2.887222519788133,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.396,
      "step": 12127
    },
    {
      "epoch": 2.8874605725168125,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 12128
    },
    {
      "epoch": 2.887698625245492,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 12129
    },
    {
      "epoch": 2.8879366779741713,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4032,
      "step": 12130
    },
    {
      "epoch": 2.8881747307028505,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1485,
      "step": 12131
    },
    {
      "epoch": 2.88841278343153,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 12132
    },
    {
      "epoch": 2.8886508361602097,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 12133
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 12134
    },
    {
      "epoch": 2.8891269416175684,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 12135
    },
    {
      "epoch": 2.8893649943462476,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 12136
    },
    {
      "epoch": 2.889603047074927,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 12137
    },
    {
      "epoch": 2.8898410998036064,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 12138
    },
    {
      "epoch": 2.890079152532286,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 12139
    },
    {
      "epoch": 2.890317205260965,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3816,
      "step": 12140
    },
    {
      "epoch": 2.8905552579896447,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 12141
    },
    {
      "epoch": 2.8907933107183243,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4379,
      "step": 12142
    },
    {
      "epoch": 2.8910313634470035,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 12143
    },
    {
      "epoch": 2.8912694161756827,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 12144
    },
    {
      "epoch": 2.8915074689043623,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 12145
    },
    {
      "epoch": 2.891745521633042,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 12146
    },
    {
      "epoch": 2.891983574361721,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3328,
      "step": 12147
    },
    {
      "epoch": 2.8922216270904006,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 12148
    },
    {
      "epoch": 2.89245967981908,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 12149
    },
    {
      "epoch": 2.8926977325477594,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3142,
      "step": 12150
    },
    {
      "epoch": 2.8929357852764386,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 12151
    },
    {
      "epoch": 2.893173838005118,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3358,
      "step": 12152
    },
    {
      "epoch": 2.8934118907337973,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 12153
    },
    {
      "epoch": 2.893649943462477,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 12154
    },
    {
      "epoch": 2.8938879961911566,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 12155
    },
    {
      "epoch": 2.8941260489198357,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 12156
    },
    {
      "epoch": 2.894364101648515,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 12157
    },
    {
      "epoch": 2.8946021543771945,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 12158
    },
    {
      "epoch": 2.894840207105874,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 12159
    },
    {
      "epoch": 2.8950782598345532,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 12160
    },
    {
      "epoch": 2.895316312563233,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3734,
      "step": 12161
    },
    {
      "epoch": 2.895554365291912,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 12162
    },
    {
      "epoch": 2.8957924180205916,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 12163
    },
    {
      "epoch": 2.8960304707492712,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 12164
    },
    {
      "epoch": 2.8962685234779504,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 12165
    },
    {
      "epoch": 2.8965065762066295,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 12166
    },
    {
      "epoch": 2.896744628935309,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 12167
    },
    {
      "epoch": 2.8969826816639888,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4436,
      "step": 12168
    },
    {
      "epoch": 2.897220734392668,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 12169
    },
    {
      "epoch": 2.8974587871213475,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4209,
      "step": 12170
    },
    {
      "epoch": 2.8976968398500267,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 12171
    },
    {
      "epoch": 2.8979348925787063,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 12172
    },
    {
      "epoch": 2.8981729453073855,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 12173
    },
    {
      "epoch": 2.898410998036065,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2749,
      "step": 12174
    },
    {
      "epoch": 2.898649050764744,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 12175
    },
    {
      "epoch": 2.898887103493424,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3122,
      "step": 12176
    },
    {
      "epoch": 2.8991251562221034,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 12177
    },
    {
      "epoch": 2.8993632089507826,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 12178
    },
    {
      "epoch": 2.8996012616794618,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 12179
    },
    {
      "epoch": 2.8998393144081414,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 12180
    },
    {
      "epoch": 2.900077367136821,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 12181
    },
    {
      "epoch": 2.9003154198655,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3682,
      "step": 12182
    },
    {
      "epoch": 2.9005534725941797,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 12183
    },
    {
      "epoch": 2.900791525322859,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 12184
    },
    {
      "epoch": 2.9010295780515385,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 12185
    },
    {
      "epoch": 2.9012676307802177,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3432,
      "step": 12186
    },
    {
      "epoch": 2.9015056835088973,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 12187
    },
    {
      "epoch": 2.9017437362375764,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 12188
    },
    {
      "epoch": 2.901981788966256,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.25,
      "step": 12189
    },
    {
      "epoch": 2.9022198416949356,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 12190
    },
    {
      "epoch": 2.902457894423615,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 12191
    },
    {
      "epoch": 2.9026959471522944,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.4159,
      "step": 12192
    },
    {
      "epoch": 2.9029339998809736,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 12193
    },
    {
      "epoch": 2.903172052609653,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 12194
    },
    {
      "epoch": 2.9034101053383323,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 12195
    },
    {
      "epoch": 2.903648158067012,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 12196
    },
    {
      "epoch": 2.903886210795691,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 12197
    },
    {
      "epoch": 2.9041242635243707,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 12198
    },
    {
      "epoch": 2.9043623162530503,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 12199
    },
    {
      "epoch": 2.9046003689817295,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 12200
    },
    {
      "epoch": 2.9046003689817295,
      "eval_runtime": 0.5142,
      "eval_samples_per_second": 1.945,
      "eval_steps_per_second": 1.945,
      "step": 12200
    },
    {
      "epoch": 2.9048384217104086,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 12201
    },
    {
      "epoch": 2.9050764744390882,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 12202
    },
    {
      "epoch": 2.905314527167768,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 12203
    },
    {
      "epoch": 2.905552579896447,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2841,
      "step": 12204
    },
    {
      "epoch": 2.9057906326251266,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 12205
    },
    {
      "epoch": 2.9060286853538058,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 12206
    },
    {
      "epoch": 2.9062667380824854,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 12207
    },
    {
      "epoch": 2.9065047908111645,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 12208
    },
    {
      "epoch": 2.906742843539844,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.382,
      "step": 12209
    },
    {
      "epoch": 2.9069808962685233,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 12210
    },
    {
      "epoch": 2.907218948997203,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 12211
    },
    {
      "epoch": 2.9074570017258825,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 12212
    },
    {
      "epoch": 2.9076950544545617,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.5249,
      "step": 12213
    },
    {
      "epoch": 2.907933107183241,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 12214
    },
    {
      "epoch": 2.9081711599119204,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 12215
    },
    {
      "epoch": 2.9084092126406,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 12216
    },
    {
      "epoch": 2.908647265369279,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4903,
      "step": 12217
    },
    {
      "epoch": 2.908885318097959,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.6397,
      "step": 12218
    },
    {
      "epoch": 2.909123370826638,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 12219
    },
    {
      "epoch": 2.9093614235553176,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 12220
    },
    {
      "epoch": 2.909599476283997,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.407,
      "step": 12221
    },
    {
      "epoch": 2.9098375290126763,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 12222
    },
    {
      "epoch": 2.9100755817413555,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 12223
    },
    {
      "epoch": 2.910313634470035,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 12224
    },
    {
      "epoch": 2.9105516871987147,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 12225
    },
    {
      "epoch": 2.910789739927394,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 12226
    },
    {
      "epoch": 2.9110277926560735,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 12227
    },
    {
      "epoch": 2.9112658453847526,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 12228
    },
    {
      "epoch": 2.9115038981134322,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 12229
    },
    {
      "epoch": 2.9117419508421114,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 12230
    },
    {
      "epoch": 2.911980003570791,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 12231
    },
    {
      "epoch": 2.91221805629947,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 12232
    },
    {
      "epoch": 2.9124561090281498,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3212,
      "step": 12233
    },
    {
      "epoch": 2.9126941617568294,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 12234
    },
    {
      "epoch": 2.9129322144855085,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 12235
    },
    {
      "epoch": 2.9131702672141877,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 12236
    },
    {
      "epoch": 2.9134083199428673,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3413,
      "step": 12237
    },
    {
      "epoch": 2.913646372671547,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 12238
    },
    {
      "epoch": 2.913884425400226,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 12239
    },
    {
      "epoch": 2.9141224781289057,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 12240
    },
    {
      "epoch": 2.914360530857585,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 12241
    },
    {
      "epoch": 2.9145985835862644,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 12242
    },
    {
      "epoch": 2.9148366363149436,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 12243
    },
    {
      "epoch": 2.915074689043623,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 12244
    },
    {
      "epoch": 2.9153127417723024,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2454,
      "step": 12245
    },
    {
      "epoch": 2.915550794500982,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 12246
    },
    {
      "epoch": 2.9157888472296616,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 12247
    },
    {
      "epoch": 2.9160268999583407,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 12248
    },
    {
      "epoch": 2.91626495268702,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.6345,
      "step": 12249
    },
    {
      "epoch": 2.9165030054156995,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3748,
      "step": 12250
    },
    {
      "epoch": 2.916741058144379,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3532,
      "step": 12251
    },
    {
      "epoch": 2.9169791108730583,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.5007,
      "step": 12252
    },
    {
      "epoch": 2.917217163601738,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 12253
    },
    {
      "epoch": 2.917455216330417,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 12254
    },
    {
      "epoch": 2.9176932690590967,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 12255
    },
    {
      "epoch": 2.9179313217877763,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 12256
    },
    {
      "epoch": 2.9181693745164554,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 12257
    },
    {
      "epoch": 2.9184074272451346,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3787,
      "step": 12258
    },
    {
      "epoch": 2.918645479973814,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 12259
    },
    {
      "epoch": 2.918883532702494,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 12260
    },
    {
      "epoch": 2.919121585431173,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 12261
    },
    {
      "epoch": 2.9193596381598526,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 12262
    },
    {
      "epoch": 2.9195976908885317,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 12263
    },
    {
      "epoch": 2.9198357436172113,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 12264
    },
    {
      "epoch": 2.9200737963458905,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 12265
    },
    {
      "epoch": 2.92031184907457,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 12266
    },
    {
      "epoch": 2.9205499018032492,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4126,
      "step": 12267
    },
    {
      "epoch": 2.920787954531929,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 12268
    },
    {
      "epoch": 2.9210260072606085,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 12269
    },
    {
      "epoch": 2.9212640599892876,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 12270
    },
    {
      "epoch": 2.921502112717967,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 12271
    },
    {
      "epoch": 2.9217401654466464,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 12272
    },
    {
      "epoch": 2.921978218175326,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 12273
    },
    {
      "epoch": 2.922216270904005,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 12274
    },
    {
      "epoch": 2.9224543236326848,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 12275
    },
    {
      "epoch": 2.922692376361364,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3943,
      "step": 12276
    },
    {
      "epoch": 2.9229304290900435,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 12277
    },
    {
      "epoch": 2.9231684818187227,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 12278
    },
    {
      "epoch": 2.9234065345474023,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 12279
    },
    {
      "epoch": 2.9236445872760815,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 12280
    },
    {
      "epoch": 2.923882640004761,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4223,
      "step": 12281
    },
    {
      "epoch": 2.9241206927334407,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3727,
      "step": 12282
    },
    {
      "epoch": 2.92435874546212,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 12283
    },
    {
      "epoch": 2.924596798190799,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 12284
    },
    {
      "epoch": 2.9248348509194786,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 12285
    },
    {
      "epoch": 2.925072903648158,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.5049,
      "step": 12286
    },
    {
      "epoch": 2.9253109563768374,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 12287
    },
    {
      "epoch": 2.925549009105517,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 12288
    },
    {
      "epoch": 2.925787061834196,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 12289
    },
    {
      "epoch": 2.9260251145628757,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4164,
      "step": 12290
    },
    {
      "epoch": 2.9262631672915553,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 12291
    },
    {
      "epoch": 2.9265012200202345,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 12292
    },
    {
      "epoch": 2.9267392727489137,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 12293
    },
    {
      "epoch": 2.9269773254775933,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 12294
    },
    {
      "epoch": 2.927215378206273,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 12295
    },
    {
      "epoch": 2.927453430934952,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3,
      "step": 12296
    },
    {
      "epoch": 2.9276914836636316,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 12297
    },
    {
      "epoch": 2.927929536392311,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 12298
    },
    {
      "epoch": 2.9281675891209904,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 12299
    },
    {
      "epoch": 2.9284056418496696,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 12300
    },
    {
      "epoch": 2.9284056418496696,
      "eval_runtime": 0.5212,
      "eval_samples_per_second": 1.919,
      "eval_steps_per_second": 1.919,
      "step": 12300
    },
    {
      "epoch": 2.928643694578349,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 12301
    },
    {
      "epoch": 2.9288817473070283,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 12302
    },
    {
      "epoch": 2.929119800035708,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 12303
    },
    {
      "epoch": 2.9293578527643875,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.472,
      "step": 12304
    },
    {
      "epoch": 2.9295959054930667,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 12305
    },
    {
      "epoch": 2.929833958221746,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3926,
      "step": 12306
    },
    {
      "epoch": 2.9300720109504255,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3246,
      "step": 12307
    },
    {
      "epoch": 2.930310063679105,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 12308
    },
    {
      "epoch": 2.9305481164077842,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 12309
    },
    {
      "epoch": 2.930786169136464,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.397,
      "step": 12310
    },
    {
      "epoch": 2.931024221865143,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 12311
    },
    {
      "epoch": 2.9312622745938226,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 12312
    },
    {
      "epoch": 2.931500327322502,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 12313
    },
    {
      "epoch": 2.9317383800511814,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 12314
    },
    {
      "epoch": 2.9319764327798605,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 12315
    },
    {
      "epoch": 2.93221448550854,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 12316
    },
    {
      "epoch": 2.9324525382372197,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.355,
      "step": 12317
    },
    {
      "epoch": 2.932690590965899,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4543,
      "step": 12318
    },
    {
      "epoch": 2.9329286436945785,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 12319
    },
    {
      "epoch": 2.9331666964232577,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3305,
      "step": 12320
    },
    {
      "epoch": 2.9334047491519373,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 12321
    },
    {
      "epoch": 2.9336428018806164,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 12322
    },
    {
      "epoch": 2.933880854609296,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 12323
    },
    {
      "epoch": 2.934118907337975,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 12324
    },
    {
      "epoch": 2.934356960066655,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 12325
    },
    {
      "epoch": 2.9345950127953344,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 12326
    },
    {
      "epoch": 2.9348330655240136,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2477,
      "step": 12327
    },
    {
      "epoch": 2.9350711182526927,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 12328
    },
    {
      "epoch": 2.9353091709813723,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 12329
    },
    {
      "epoch": 2.935547223710052,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 12330
    },
    {
      "epoch": 2.935785276438731,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 12331
    },
    {
      "epoch": 2.9360233291674107,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 12332
    },
    {
      "epoch": 2.93626138189609,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 12333
    },
    {
      "epoch": 2.9364994346247695,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 12334
    },
    {
      "epoch": 2.9367374873534486,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 12335
    },
    {
      "epoch": 2.9369755400821282,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 12336
    },
    {
      "epoch": 2.9372135928108074,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5153,
      "step": 12337
    },
    {
      "epoch": 2.937451645539487,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 12338
    },
    {
      "epoch": 2.9376896982681666,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3992,
      "step": 12339
    },
    {
      "epoch": 2.9379277509968458,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3489,
      "step": 12340
    },
    {
      "epoch": 2.938165803725525,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 12341
    },
    {
      "epoch": 2.9384038564542045,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 12342
    },
    {
      "epoch": 2.938641909182884,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 12343
    },
    {
      "epoch": 2.9388799619115633,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 12344
    },
    {
      "epoch": 2.939118014640243,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 12345
    },
    {
      "epoch": 2.939356067368922,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 12346
    },
    {
      "epoch": 2.9395941200976017,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 12347
    },
    {
      "epoch": 2.9398321728262813,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 12348
    },
    {
      "epoch": 2.9400702255549604,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2882,
      "step": 12349
    },
    {
      "epoch": 2.9403082782836396,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 12350
    },
    {
      "epoch": 2.940546331012319,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.6442,
      "step": 12351
    },
    {
      "epoch": 2.940784383740999,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 12352
    },
    {
      "epoch": 2.941022436469678,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.316,
      "step": 12353
    },
    {
      "epoch": 2.9412604891983576,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 12354
    },
    {
      "epoch": 2.9414985419270367,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 12355
    },
    {
      "epoch": 2.9417365946557164,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2977,
      "step": 12356
    },
    {
      "epoch": 2.9419746473843955,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3293,
      "step": 12357
    },
    {
      "epoch": 2.942212700113075,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3578,
      "step": 12358
    },
    {
      "epoch": 2.9424507528417543,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 12359
    },
    {
      "epoch": 2.942688805570434,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 12360
    },
    {
      "epoch": 2.9429268582991135,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 12361
    },
    {
      "epoch": 2.9431649110277927,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 12362
    },
    {
      "epoch": 2.943402963756472,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3467,
      "step": 12363
    },
    {
      "epoch": 2.9436410164851514,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 12364
    },
    {
      "epoch": 2.943879069213831,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 12365
    },
    {
      "epoch": 2.94411712194251,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 12366
    },
    {
      "epoch": 2.94435517467119,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 12367
    },
    {
      "epoch": 2.944593227399869,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 12368
    },
    {
      "epoch": 2.9448312801285486,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 12369
    },
    {
      "epoch": 2.9450693328572277,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3988,
      "step": 12370
    },
    {
      "epoch": 2.9453073855859073,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 12371
    },
    {
      "epoch": 2.9455454383145865,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.373,
      "step": 12372
    },
    {
      "epoch": 2.945783491043266,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 12373
    },
    {
      "epoch": 2.9460215437719457,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 12374
    },
    {
      "epoch": 2.946259596500625,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 12375
    },
    {
      "epoch": 2.946497649229304,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 12376
    },
    {
      "epoch": 2.9467357019579836,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 12377
    },
    {
      "epoch": 2.9469737546866632,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 12378
    },
    {
      "epoch": 2.9472118074153424,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 12379
    },
    {
      "epoch": 2.947449860144022,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3566,
      "step": 12380
    },
    {
      "epoch": 2.947687912872701,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 12381
    },
    {
      "epoch": 2.9479259656013808,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 12382
    },
    {
      "epoch": 2.9481640183300604,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 12383
    },
    {
      "epoch": 2.9484020710587395,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3942,
      "step": 12384
    },
    {
      "epoch": 2.9486401237874187,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 12385
    },
    {
      "epoch": 2.9488781765160983,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 12386
    },
    {
      "epoch": 2.949116229244778,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 12387
    },
    {
      "epoch": 2.949354281973457,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 12388
    },
    {
      "epoch": 2.9495923347021367,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 12389
    },
    {
      "epoch": 2.949830387430816,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 12390
    },
    {
      "epoch": 2.9500684401594954,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 12391
    },
    {
      "epoch": 2.9503064928881746,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 12392
    },
    {
      "epoch": 2.950544545616854,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3997,
      "step": 12393
    },
    {
      "epoch": 2.9507825983455334,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 12394
    },
    {
      "epoch": 2.951020651074213,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 12395
    },
    {
      "epoch": 2.9512587038028926,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 12396
    },
    {
      "epoch": 2.9514967565315717,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 12397
    },
    {
      "epoch": 2.951734809260251,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 12398
    },
    {
      "epoch": 2.9519728619889305,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 12399
    },
    {
      "epoch": 2.95221091471761,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 12400
    },
    {
      "epoch": 2.95221091471761,
      "eval_runtime": 0.5167,
      "eval_samples_per_second": 1.935,
      "eval_steps_per_second": 1.935,
      "step": 12400
    },
    {
      "epoch": 2.9524489674462893,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3443,
      "step": 12401
    },
    {
      "epoch": 2.952687020174969,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4581,
      "step": 12402
    },
    {
      "epoch": 2.952925072903648,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 12403
    },
    {
      "epoch": 2.9531631256323276,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 12404
    },
    {
      "epoch": 2.9534011783610072,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 12405
    },
    {
      "epoch": 2.9536392310896864,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3138,
      "step": 12406
    },
    {
      "epoch": 2.9538772838183656,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.6803,
      "step": 12407
    },
    {
      "epoch": 2.954115336547045,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 12408
    },
    {
      "epoch": 2.9543533892757248,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 12409
    },
    {
      "epoch": 2.954591442004404,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 12410
    },
    {
      "epoch": 2.9548294947330835,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 12411
    },
    {
      "epoch": 2.9550675474617627,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 12412
    },
    {
      "epoch": 2.9553056001904423,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 12413
    },
    {
      "epoch": 2.9555436529191215,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 12414
    },
    {
      "epoch": 2.955781705647801,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3843,
      "step": 12415
    },
    {
      "epoch": 2.9560197583764802,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 12416
    },
    {
      "epoch": 2.95625781110516,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 12417
    },
    {
      "epoch": 2.9564958638338394,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 12418
    },
    {
      "epoch": 2.9567339165625186,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 12419
    },
    {
      "epoch": 2.9569719692911978,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 12420
    },
    {
      "epoch": 2.9572100220198774,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3616,
      "step": 12421
    },
    {
      "epoch": 2.957448074748557,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 12422
    },
    {
      "epoch": 2.957686127477236,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 12423
    },
    {
      "epoch": 2.9579241802059157,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 12424
    },
    {
      "epoch": 2.958162232934595,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 12425
    },
    {
      "epoch": 2.9584002856632745,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 12426
    },
    {
      "epoch": 2.9586383383919537,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 12427
    },
    {
      "epoch": 2.9588763911206333,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 12428
    },
    {
      "epoch": 2.9591144438493124,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 12429
    },
    {
      "epoch": 2.959352496577992,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 12430
    },
    {
      "epoch": 2.9595905493066716,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3527,
      "step": 12431
    },
    {
      "epoch": 2.959828602035351,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 12432
    },
    {
      "epoch": 2.96006665476403,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 12433
    },
    {
      "epoch": 2.9603047074927096,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3542,
      "step": 12434
    },
    {
      "epoch": 2.960542760221389,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.4219,
      "step": 12435
    },
    {
      "epoch": 2.9607808129500683,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 12436
    },
    {
      "epoch": 2.961018865678748,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 12437
    },
    {
      "epoch": 2.961256918407427,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 12438
    },
    {
      "epoch": 2.9614949711361067,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 12439
    },
    {
      "epoch": 2.9617330238647863,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.339,
      "step": 12440
    },
    {
      "epoch": 2.9619710765934655,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 12441
    },
    {
      "epoch": 2.9622091293221446,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 12442
    },
    {
      "epoch": 2.9624471820508242,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 12443
    },
    {
      "epoch": 2.962685234779504,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 12444
    },
    {
      "epoch": 2.962923287508183,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 12445
    },
    {
      "epoch": 2.9631613402368626,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 12446
    },
    {
      "epoch": 2.963399392965542,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 12447
    },
    {
      "epoch": 2.9636374456942214,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 12448
    },
    {
      "epoch": 2.9638754984229005,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3813,
      "step": 12449
    },
    {
      "epoch": 2.96411355115158,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3557,
      "step": 12450
    },
    {
      "epoch": 2.9643516038802593,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 12451
    },
    {
      "epoch": 2.964589656608939,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 12452
    },
    {
      "epoch": 2.9648277093376185,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 12453
    },
    {
      "epoch": 2.9650657620662977,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 12454
    },
    {
      "epoch": 2.965303814794977,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 12455
    },
    {
      "epoch": 2.9655418675236564,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 12456
    },
    {
      "epoch": 2.965779920252336,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 12457
    },
    {
      "epoch": 2.966017972981015,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 12458
    },
    {
      "epoch": 2.966256025709695,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 12459
    },
    {
      "epoch": 2.966494078438374,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 12460
    },
    {
      "epoch": 2.9667321311670536,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 12461
    },
    {
      "epoch": 2.9669701838957327,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 12462
    },
    {
      "epoch": 2.9672082366244124,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 12463
    },
    {
      "epoch": 2.9674462893530915,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 12464
    },
    {
      "epoch": 2.967684342081771,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 12465
    },
    {
      "epoch": 2.9679223948104507,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 12466
    },
    {
      "epoch": 2.96816044753913,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 12467
    },
    {
      "epoch": 2.968398500267809,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 12468
    },
    {
      "epoch": 2.9686365529964887,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2785,
      "step": 12469
    },
    {
      "epoch": 2.9688746057251683,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4086,
      "step": 12470
    },
    {
      "epoch": 2.9691126584538474,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 12471
    },
    {
      "epoch": 2.969350711182527,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.6012,
      "step": 12472
    },
    {
      "epoch": 2.969588763911206,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 12473
    },
    {
      "epoch": 2.969826816639886,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 12474
    },
    {
      "epoch": 2.9700648693685654,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7966,
      "step": 12475
    },
    {
      "epoch": 2.9703029220972446,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 12476
    },
    {
      "epoch": 2.9705409748259237,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 12477
    },
    {
      "epoch": 2.9707790275546033,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 12478
    },
    {
      "epoch": 2.971017080283283,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4148,
      "step": 12479
    },
    {
      "epoch": 2.971255133011962,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 12480
    },
    {
      "epoch": 2.9714931857406417,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 12481
    },
    {
      "epoch": 2.971731238469321,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 12482
    },
    {
      "epoch": 2.9719692911980005,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 12483
    },
    {
      "epoch": 2.9722073439266796,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 12484
    },
    {
      "epoch": 2.9724453966553592,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 12485
    },
    {
      "epoch": 2.9726834493840384,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3683,
      "step": 12486
    },
    {
      "epoch": 2.972921502112718,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 12487
    },
    {
      "epoch": 2.9731595548413976,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 12488
    },
    {
      "epoch": 2.9733976075700768,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 12489
    },
    {
      "epoch": 2.973635660298756,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 12490
    },
    {
      "epoch": 2.9738737130274355,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2883,
      "step": 12491
    },
    {
      "epoch": 2.974111765756115,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.4678,
      "step": 12492
    },
    {
      "epoch": 2.9743498184847943,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 12493
    },
    {
      "epoch": 2.974587871213474,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 12494
    },
    {
      "epoch": 2.974825923942153,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 12495
    },
    {
      "epoch": 2.9750639766708327,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 12496
    },
    {
      "epoch": 2.975302029399512,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 12497
    },
    {
      "epoch": 2.9755400821281914,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 12498
    },
    {
      "epoch": 2.9757781348568706,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3812,
      "step": 12499
    },
    {
      "epoch": 2.97601618758555,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 12500
    },
    {
      "epoch": 2.97601618758555,
      "eval_runtime": 0.5147,
      "eval_samples_per_second": 1.943,
      "eval_steps_per_second": 1.943,
      "step": 12500
    },
    {
      "epoch": 2.97625424031423,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 12501
    },
    {
      "epoch": 2.976492293042909,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3717,
      "step": 12502
    },
    {
      "epoch": 2.9767303457715886,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 12503
    },
    {
      "epoch": 2.9769683985002677,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 12504
    },
    {
      "epoch": 2.9772064512289473,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3491,
      "step": 12505
    },
    {
      "epoch": 2.9774445039576265,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3338,
      "step": 12506
    },
    {
      "epoch": 2.977682556686306,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 12507
    },
    {
      "epoch": 2.9779206094149853,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 12508
    },
    {
      "epoch": 2.978158662143665,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3319,
      "step": 12509
    },
    {
      "epoch": 2.9783967148723445,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 12510
    },
    {
      "epoch": 2.9786347676010236,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 12511
    },
    {
      "epoch": 2.978872820329703,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 12512
    },
    {
      "epoch": 2.9791108730583824,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 12513
    },
    {
      "epoch": 2.979348925787062,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.5033,
      "step": 12514
    },
    {
      "epoch": 2.979586978515741,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 12515
    },
    {
      "epoch": 2.9798250312444208,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 12516
    },
    {
      "epoch": 2.9800630839731,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 12517
    },
    {
      "epoch": 2.9803011367017795,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 12518
    },
    {
      "epoch": 2.9805391894304587,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 12519
    },
    {
      "epoch": 2.9807772421591383,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 12520
    },
    {
      "epoch": 2.9810152948878175,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3052,
      "step": 12521
    },
    {
      "epoch": 2.981253347616497,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 12522
    },
    {
      "epoch": 2.9814914003451767,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.365,
      "step": 12523
    },
    {
      "epoch": 2.981729453073856,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5104,
      "step": 12524
    },
    {
      "epoch": 2.981967505802535,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 12525
    },
    {
      "epoch": 2.9822055585312146,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2819,
      "step": 12526
    },
    {
      "epoch": 2.982443611259894,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 12527
    },
    {
      "epoch": 2.9826816639885734,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 12528
    },
    {
      "epoch": 2.982919716717253,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 12529
    },
    {
      "epoch": 2.983157769445932,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 12530
    },
    {
      "epoch": 2.9833958221746117,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 12531
    },
    {
      "epoch": 2.9836338749032913,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2591,
      "step": 12532
    },
    {
      "epoch": 2.9838719276319705,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 12533
    },
    {
      "epoch": 2.9841099803606497,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 12534
    },
    {
      "epoch": 2.9843480330893293,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4295,
      "step": 12535
    },
    {
      "epoch": 2.984586085818009,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7323,
      "step": 12536
    },
    {
      "epoch": 2.984824138546688,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 12537
    },
    {
      "epoch": 2.9850621912753676,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 12538
    },
    {
      "epoch": 2.985300244004047,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4553,
      "step": 12539
    },
    {
      "epoch": 2.9855382967327264,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 12540
    },
    {
      "epoch": 2.9857763494614056,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 12541
    },
    {
      "epoch": 2.986014402190085,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 12542
    },
    {
      "epoch": 2.9862524549187643,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7061,
      "step": 12543
    },
    {
      "epoch": 2.986490507647444,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 12544
    },
    {
      "epoch": 2.9867285603761236,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 12545
    },
    {
      "epoch": 2.9869666131048027,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3701,
      "step": 12546
    },
    {
      "epoch": 2.987204665833482,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2183,
      "step": 12547
    },
    {
      "epoch": 2.9874427185621615,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 12548
    },
    {
      "epoch": 2.987680771290841,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 12549
    },
    {
      "epoch": 2.9879188240195202,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 12550
    },
    {
      "epoch": 2.9881568767482,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 12551
    },
    {
      "epoch": 2.988394929476879,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 12552
    },
    {
      "epoch": 2.9886329822055586,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 12553
    },
    {
      "epoch": 2.988871034934238,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 12554
    },
    {
      "epoch": 2.9891090876629174,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 12555
    },
    {
      "epoch": 2.9893471403915965,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2835,
      "step": 12556
    },
    {
      "epoch": 2.989585193120276,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 12557
    },
    {
      "epoch": 2.9898232458489558,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 12558
    },
    {
      "epoch": 2.990061298577635,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 12559
    },
    {
      "epoch": 2.990299351306314,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 12560
    },
    {
      "epoch": 2.9905374040349937,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 12561
    },
    {
      "epoch": 2.9907754567636733,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 12562
    },
    {
      "epoch": 2.9910135094923525,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 12563
    },
    {
      "epoch": 2.991251562221032,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 12564
    },
    {
      "epoch": 2.991489614949711,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3486,
      "step": 12565
    },
    {
      "epoch": 2.991727667678391,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 12566
    },
    {
      "epoch": 2.9919657204070704,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.458,
      "step": 12567
    },
    {
      "epoch": 2.9922037731357496,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 12568
    },
    {
      "epoch": 2.9924418258644288,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 12569
    },
    {
      "epoch": 2.9926798785931084,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 12570
    },
    {
      "epoch": 2.992917931321788,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 12571
    },
    {
      "epoch": 2.993155984050467,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 12572
    },
    {
      "epoch": 2.9933940367791467,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 12573
    },
    {
      "epoch": 2.993632089507826,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 12574
    },
    {
      "epoch": 2.9938701422365055,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 12575
    },
    {
      "epoch": 2.9941081949651847,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 12576
    },
    {
      "epoch": 2.9943462476938643,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 12577
    },
    {
      "epoch": 2.9945843004225434,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 12578
    },
    {
      "epoch": 2.994822353151223,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 12579
    },
    {
      "epoch": 2.9950604058799026,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 12580
    },
    {
      "epoch": 2.995298458608582,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 12581
    },
    {
      "epoch": 2.995536511337261,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 12582
    },
    {
      "epoch": 2.9957745640659406,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4555,
      "step": 12583
    },
    {
      "epoch": 2.99601261679462,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 12584
    },
    {
      "epoch": 2.9962506695232993,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 12585
    },
    {
      "epoch": 2.996488722251979,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 12586
    },
    {
      "epoch": 2.996726774980658,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 12587
    },
    {
      "epoch": 2.9969648277093377,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 12588
    },
    {
      "epoch": 2.997202880438017,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4481,
      "step": 12589
    },
    {
      "epoch": 2.9974409331666965,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 12590
    },
    {
      "epoch": 2.9976789858953756,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 12591
    },
    {
      "epoch": 2.9979170386240552,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 12592
    },
    {
      "epoch": 2.998155091352735,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 12593
    },
    {
      "epoch": 2.998393144081414,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 12594
    },
    {
      "epoch": 2.9986311968100936,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4227,
      "step": 12595
    },
    {
      "epoch": 2.9988692495387728,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 12596
    },
    {
      "epoch": 2.9991073022674524,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 12597
    },
    {
      "epoch": 2.9993453549961315,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 12598
    },
    {
      "epoch": 2.999583407724811,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 12599
    },
    {
      "epoch": 2.9998214604534903,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 12600
    },
    {
      "epoch": 2.9998214604534903,
      "eval_runtime": 0.5083,
      "eval_samples_per_second": 1.967,
      "eval_steps_per_second": 1.967,
      "step": 12600
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 12601
    },
    {
      "epoch": 3.0002380527286796,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 12602
    },
    {
      "epoch": 3.0004761054573588,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 12603
    },
    {
      "epoch": 3.0007141581860384,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 12604
    },
    {
      "epoch": 3.0009522109147175,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 12605
    },
    {
      "epoch": 3.001190263643397,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 12606
    },
    {
      "epoch": 3.0014283163720763,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 12607
    },
    {
      "epoch": 3.001666369100756,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 12608
    },
    {
      "epoch": 3.001904421829435,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 12609
    },
    {
      "epoch": 3.0021424745581147,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 12610
    },
    {
      "epoch": 3.002380527286794,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 12611
    },
    {
      "epoch": 3.0026185800154734,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 12612
    },
    {
      "epoch": 3.002856632744153,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3207,
      "step": 12613
    },
    {
      "epoch": 3.003094685472832,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 12614
    },
    {
      "epoch": 3.003332738201512,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 12615
    },
    {
      "epoch": 3.003570790930191,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 12616
    },
    {
      "epoch": 3.0038088436588706,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 12617
    },
    {
      "epoch": 3.0040468963875497,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 12618
    },
    {
      "epoch": 3.0042849491162293,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 12619
    },
    {
      "epoch": 3.0045230018449085,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 12620
    },
    {
      "epoch": 3.004761054573588,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3473,
      "step": 12621
    },
    {
      "epoch": 3.0049991073022673,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 12622
    },
    {
      "epoch": 3.005237160030947,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 12623
    },
    {
      "epoch": 3.005475212759626,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3497,
      "step": 12624
    },
    {
      "epoch": 3.0057132654883056,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 12625
    },
    {
      "epoch": 3.0059513182169852,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 12626
    },
    {
      "epoch": 3.0061893709456644,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 12627
    },
    {
      "epoch": 3.006427423674344,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 12628
    },
    {
      "epoch": 3.006665476403023,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 12629
    },
    {
      "epoch": 3.006903529131703,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 12630
    },
    {
      "epoch": 3.007141581860382,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 12631
    },
    {
      "epoch": 3.0073796345890615,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 12632
    },
    {
      "epoch": 3.0076176873177407,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 12633
    },
    {
      "epoch": 3.0078557400464203,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 12634
    },
    {
      "epoch": 3.0080937927750995,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 12635
    },
    {
      "epoch": 3.008331845503779,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 12636
    },
    {
      "epoch": 3.0085698982324587,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 12637
    },
    {
      "epoch": 3.008807950961138,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 12638
    },
    {
      "epoch": 3.0090460036898175,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1368,
      "step": 12639
    },
    {
      "epoch": 3.0092840564184966,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 12640
    },
    {
      "epoch": 3.009522109147176,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3805,
      "step": 12641
    },
    {
      "epoch": 3.0097601618758554,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 12642
    },
    {
      "epoch": 3.009998214604535,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 12643
    },
    {
      "epoch": 3.010236267333214,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 12644
    },
    {
      "epoch": 3.0104743200618937,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 12645
    },
    {
      "epoch": 3.010712372790573,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 12646
    },
    {
      "epoch": 3.0109504255192525,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4211,
      "step": 12647
    },
    {
      "epoch": 3.011188478247932,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 12648
    },
    {
      "epoch": 3.0114265309766113,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 12649
    },
    {
      "epoch": 3.011664583705291,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 12650
    },
    {
      "epoch": 3.01190263643397,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 12651
    },
    {
      "epoch": 3.0121406891626497,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 12652
    },
    {
      "epoch": 3.012378741891329,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 12653
    },
    {
      "epoch": 3.0126167946200084,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 12654
    },
    {
      "epoch": 3.0128548473486876,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 12655
    },
    {
      "epoch": 3.013092900077367,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 12656
    },
    {
      "epoch": 3.0133309528060463,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 12657
    },
    {
      "epoch": 3.013569005534726,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 12658
    },
    {
      "epoch": 3.0138070582634056,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9622,
      "step": 12659
    },
    {
      "epoch": 3.0140451109920847,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 12660
    },
    {
      "epoch": 3.0142831637207643,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 12661
    },
    {
      "epoch": 3.0145212164494435,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 12662
    },
    {
      "epoch": 3.014759269178123,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 12663
    },
    {
      "epoch": 3.0149973219068023,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 12664
    },
    {
      "epoch": 3.015235374635482,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 12665
    },
    {
      "epoch": 3.015473427364161,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 12666
    },
    {
      "epoch": 3.0157114800928406,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 12667
    },
    {
      "epoch": 3.01594953282152,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 12668
    },
    {
      "epoch": 3.0161875855501994,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 12669
    },
    {
      "epoch": 3.0164256382788786,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 12670
    },
    {
      "epoch": 3.016663691007558,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 12671
    },
    {
      "epoch": 3.0169017437362378,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 12672
    },
    {
      "epoch": 3.017139796464917,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 12673
    },
    {
      "epoch": 3.0173778491935965,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 12674
    },
    {
      "epoch": 3.0176159019222757,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 12675
    },
    {
      "epoch": 3.0178539546509553,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 12676
    },
    {
      "epoch": 3.0180920073796345,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 12677
    },
    {
      "epoch": 3.018330060108314,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2197,
      "step": 12678
    },
    {
      "epoch": 3.0185681128369932,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 12679
    },
    {
      "epoch": 3.018806165565673,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 12680
    },
    {
      "epoch": 3.019044218294352,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 12681
    },
    {
      "epoch": 3.0192822710230316,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 12682
    },
    {
      "epoch": 3.019520323751711,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 12683
    },
    {
      "epoch": 3.0197583764803904,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 12684
    },
    {
      "epoch": 3.01999642920907,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 12685
    },
    {
      "epoch": 3.020234481937749,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4056,
      "step": 12686
    },
    {
      "epoch": 3.0204725346664287,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3297,
      "step": 12687
    },
    {
      "epoch": 3.020710587395108,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 12688
    },
    {
      "epoch": 3.0209486401237875,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 12689
    },
    {
      "epoch": 3.0211866928524667,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 12690
    },
    {
      "epoch": 3.0214247455811463,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3719,
      "step": 12691
    },
    {
      "epoch": 3.0216627983098254,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 12692
    },
    {
      "epoch": 3.021900851038505,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 12693
    },
    {
      "epoch": 3.0221389037671846,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 12694
    },
    {
      "epoch": 3.022376956495864,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 12695
    },
    {
      "epoch": 3.0226150092245434,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 12696
    },
    {
      "epoch": 3.0228530619532226,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 12697
    },
    {
      "epoch": 3.023091114681902,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 12698
    },
    {
      "epoch": 3.0233291674105813,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 12699
    },
    {
      "epoch": 3.023567220139261,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 12700
    },
    {
      "epoch": 3.023567220139261,
      "eval_runtime": 0.5245,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 12700
    },
    {
      "epoch": 3.02380527286794,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.3617,
      "step": 12701
    },
    {
      "epoch": 3.0240433255966197,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 12702
    },
    {
      "epoch": 3.024281378325299,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 12703
    },
    {
      "epoch": 3.0245194310539785,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 12704
    },
    {
      "epoch": 3.024757483782658,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 12705
    },
    {
      "epoch": 3.0249955365113372,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 12706
    },
    {
      "epoch": 3.025233589240017,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 12707
    },
    {
      "epoch": 3.025471641968696,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 12708
    },
    {
      "epoch": 3.0257096946973756,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 12709
    },
    {
      "epoch": 3.0259477474260548,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 12710
    },
    {
      "epoch": 3.0261858001547344,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 12711
    },
    {
      "epoch": 3.0264238528834135,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 12712
    },
    {
      "epoch": 3.026661905612093,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 12713
    },
    {
      "epoch": 3.0268999583407723,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 12714
    },
    {
      "epoch": 3.027138011069452,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 12715
    },
    {
      "epoch": 3.027376063798131,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 12716
    },
    {
      "epoch": 3.0276141165268107,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3509,
      "step": 12717
    },
    {
      "epoch": 3.0278521692554903,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 12718
    },
    {
      "epoch": 3.0280902219841694,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 12719
    },
    {
      "epoch": 3.028328274712849,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8356,
      "step": 12720
    },
    {
      "epoch": 3.028566327441528,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 12721
    },
    {
      "epoch": 3.028804380170208,
      "grad_norm": 7.15625,
      "learning_rate": 3e-05,
      "loss": 2.43,
      "step": 12722
    },
    {
      "epoch": 3.029042432898887,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.3771,
      "step": 12723
    },
    {
      "epoch": 3.0292804856275666,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 12724
    },
    {
      "epoch": 3.0295185383562457,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3823,
      "step": 12725
    },
    {
      "epoch": 3.0297565910849253,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 12726
    },
    {
      "epoch": 3.0299946438136045,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 12727
    },
    {
      "epoch": 3.030232696542284,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 12728
    },
    {
      "epoch": 3.0304707492709637,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 12729
    },
    {
      "epoch": 3.030708801999643,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 12730
    },
    {
      "epoch": 3.0309468547283225,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 12731
    },
    {
      "epoch": 3.0311849074570016,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 12732
    },
    {
      "epoch": 3.0314229601856812,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 12733
    },
    {
      "epoch": 3.0316610129143604,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.4894,
      "step": 12734
    },
    {
      "epoch": 3.03189906564304,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 12735
    },
    {
      "epoch": 3.032137118371719,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 12736
    },
    {
      "epoch": 3.032375171100399,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 12737
    },
    {
      "epoch": 3.032613223829078,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.284,
      "step": 12738
    },
    {
      "epoch": 3.0328512765577575,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 12739
    },
    {
      "epoch": 3.033089329286437,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 12740
    },
    {
      "epoch": 3.0333273820151163,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 12741
    },
    {
      "epoch": 3.033565434743796,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 12742
    },
    {
      "epoch": 3.033803487472475,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 12743
    },
    {
      "epoch": 3.0340415402011547,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 12744
    },
    {
      "epoch": 3.034279592929834,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3799,
      "step": 12745
    },
    {
      "epoch": 3.0345176456585135,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 12746
    },
    {
      "epoch": 3.0347556983871926,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 12747
    },
    {
      "epoch": 3.034993751115872,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 12748
    },
    {
      "epoch": 3.0352318038445514,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 12749
    },
    {
      "epoch": 3.035469856573231,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 12750
    },
    {
      "epoch": 3.0357079093019106,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 12751
    },
    {
      "epoch": 3.0359459620305898,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 12752
    },
    {
      "epoch": 3.0361840147592694,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 12753
    },
    {
      "epoch": 3.0364220674879485,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 12754
    },
    {
      "epoch": 3.036660120216628,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 12755
    },
    {
      "epoch": 3.0368981729453073,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 12756
    },
    {
      "epoch": 3.037136225673987,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 12757
    },
    {
      "epoch": 3.037374278402666,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 12758
    },
    {
      "epoch": 3.0376123311313457,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 12759
    },
    {
      "epoch": 3.037850383860025,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 12760
    },
    {
      "epoch": 3.0380884365887044,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 12761
    },
    {
      "epoch": 3.0383264893173836,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 12762
    },
    {
      "epoch": 3.038564542046063,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 12763
    },
    {
      "epoch": 3.038802594774743,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 12764
    },
    {
      "epoch": 3.039040647503422,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 12765
    },
    {
      "epoch": 3.0392787002321016,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 12766
    },
    {
      "epoch": 3.0395167529607807,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 12767
    },
    {
      "epoch": 3.0397548056894603,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 12768
    },
    {
      "epoch": 3.0399928584181395,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 12769
    },
    {
      "epoch": 3.040230911146819,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 12770
    },
    {
      "epoch": 3.0404689638754983,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 12771
    },
    {
      "epoch": 3.040707016604178,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3735,
      "step": 12772
    },
    {
      "epoch": 3.040945069332857,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 12773
    },
    {
      "epoch": 3.0411831220615366,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 12774
    },
    {
      "epoch": 3.0414211747902162,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 12775
    },
    {
      "epoch": 3.0416592275188954,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 12776
    },
    {
      "epoch": 3.041897280247575,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 12777
    },
    {
      "epoch": 3.042135332976254,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 12778
    },
    {
      "epoch": 3.0423733857049338,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 12779
    },
    {
      "epoch": 3.042611438433613,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 12780
    },
    {
      "epoch": 3.0428494911622925,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 12781
    },
    {
      "epoch": 3.0430875438909717,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 12782
    },
    {
      "epoch": 3.0433255966196513,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3395,
      "step": 12783
    },
    {
      "epoch": 3.0435636493483305,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 12784
    },
    {
      "epoch": 3.04380170207701,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2842,
      "step": 12785
    },
    {
      "epoch": 3.0440397548056897,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 12786
    },
    {
      "epoch": 3.044277807534369,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 12787
    },
    {
      "epoch": 3.0445158602630484,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 12788
    },
    {
      "epoch": 3.0447539129917276,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 12789
    },
    {
      "epoch": 3.044991965720407,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 12790
    },
    {
      "epoch": 3.0452300184490864,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2808,
      "step": 12791
    },
    {
      "epoch": 3.045468071177766,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 12792
    },
    {
      "epoch": 3.045706123906445,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 12793
    },
    {
      "epoch": 3.0459441766351247,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 12794
    },
    {
      "epoch": 3.046182229363804,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 12795
    },
    {
      "epoch": 3.0464202820924835,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5185,
      "step": 12796
    },
    {
      "epoch": 3.046658334821163,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 12797
    },
    {
      "epoch": 3.0468963875498423,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8816,
      "step": 12798
    },
    {
      "epoch": 3.047134440278522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 12799
    },
    {
      "epoch": 3.047372493007201,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 12800
    },
    {
      "epoch": 3.047372493007201,
      "eval_runtime": 0.5095,
      "eval_samples_per_second": 1.963,
      "eval_steps_per_second": 1.963,
      "step": 12800
    },
    {
      "epoch": 3.0476105457358806,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 12801
    },
    {
      "epoch": 3.04784859846456,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 12802
    },
    {
      "epoch": 3.0480866511932394,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 12803
    },
    {
      "epoch": 3.0483247039219186,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 12804
    },
    {
      "epoch": 3.048562756650598,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3477,
      "step": 12805
    },
    {
      "epoch": 3.0488008093792773,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 12806
    },
    {
      "epoch": 3.049038862107957,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 12807
    },
    {
      "epoch": 3.049276914836636,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 12808
    },
    {
      "epoch": 3.0495149675653157,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 12809
    },
    {
      "epoch": 3.0497530202939953,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 12810
    },
    {
      "epoch": 3.0499910730226745,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 12811
    },
    {
      "epoch": 3.050229125751354,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 12812
    },
    {
      "epoch": 3.0504671784800332,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 12813
    },
    {
      "epoch": 3.050705231208713,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 12814
    },
    {
      "epoch": 3.050943283937392,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 12815
    },
    {
      "epoch": 3.0511813366660716,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4196,
      "step": 12816
    },
    {
      "epoch": 3.0514193893947508,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 12817
    },
    {
      "epoch": 3.0516574421234304,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 12818
    },
    {
      "epoch": 3.0518954948521095,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 12819
    },
    {
      "epoch": 3.052133547580789,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 12820
    },
    {
      "epoch": 3.0523716003094687,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 12821
    },
    {
      "epoch": 3.052609653038148,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 12822
    },
    {
      "epoch": 3.0528477057668275,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3593,
      "step": 12823
    },
    {
      "epoch": 3.0530857584955067,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.4607,
      "step": 12824
    },
    {
      "epoch": 3.0533238112241863,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 12825
    },
    {
      "epoch": 3.0535618639528654,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 12826
    },
    {
      "epoch": 3.053799916681545,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 12827
    },
    {
      "epoch": 3.054037969410224,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4205,
      "step": 12828
    },
    {
      "epoch": 3.054276022138904,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 12829
    },
    {
      "epoch": 3.054514074867583,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 12830
    },
    {
      "epoch": 3.0547521275962626,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 12831
    },
    {
      "epoch": 3.054990180324942,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 12832
    },
    {
      "epoch": 3.0552282330536213,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 12833
    },
    {
      "epoch": 3.055466285782301,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 12834
    },
    {
      "epoch": 3.05570433851098,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 12835
    },
    {
      "epoch": 3.0559423912396597,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 12836
    },
    {
      "epoch": 3.056180443968339,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 12837
    },
    {
      "epoch": 3.0564184966970185,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 12838
    },
    {
      "epoch": 3.0566565494256976,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 12839
    },
    {
      "epoch": 3.0568946021543772,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 12840
    },
    {
      "epoch": 3.0571326548830564,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 12841
    },
    {
      "epoch": 3.057370707611736,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 12842
    },
    {
      "epoch": 3.0576087603404156,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 12843
    },
    {
      "epoch": 3.057846813069095,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 12844
    },
    {
      "epoch": 3.0580848657977744,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 12845
    },
    {
      "epoch": 3.0583229185264535,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 12846
    },
    {
      "epoch": 3.058560971255133,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 12847
    },
    {
      "epoch": 3.0587990239838123,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3497,
      "step": 12848
    },
    {
      "epoch": 3.059037076712492,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 12849
    },
    {
      "epoch": 3.059275129441171,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3523,
      "step": 12850
    },
    {
      "epoch": 3.0595131821698507,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 12851
    },
    {
      "epoch": 3.05975123489853,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 12852
    },
    {
      "epoch": 3.0599892876272095,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 12853
    },
    {
      "epoch": 3.0602273403558886,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 12854
    },
    {
      "epoch": 3.060465393084568,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3469,
      "step": 12855
    },
    {
      "epoch": 3.060703445813248,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 12856
    },
    {
      "epoch": 3.060941498541927,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 12857
    },
    {
      "epoch": 3.0611795512706066,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 12858
    },
    {
      "epoch": 3.0614176039992858,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 12859
    },
    {
      "epoch": 3.0616556567279654,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 12860
    },
    {
      "epoch": 3.0618937094566445,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 12861
    },
    {
      "epoch": 3.062131762185324,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 12862
    },
    {
      "epoch": 3.0623698149140033,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 12863
    },
    {
      "epoch": 3.062607867642683,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 12864
    },
    {
      "epoch": 3.062845920371362,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 12865
    },
    {
      "epoch": 3.0630839731000417,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 12866
    },
    {
      "epoch": 3.0633220258287213,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 12867
    },
    {
      "epoch": 3.0635600785574004,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 12868
    },
    {
      "epoch": 3.06379813128608,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 12869
    },
    {
      "epoch": 3.064036184014759,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5671,
      "step": 12870
    },
    {
      "epoch": 3.064274236743439,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 12871
    },
    {
      "epoch": 3.064512289472118,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 12872
    },
    {
      "epoch": 3.0647503422007976,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 12873
    },
    {
      "epoch": 3.0649883949294767,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.4215,
      "step": 12874
    },
    {
      "epoch": 3.0652264476581563,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.395,
      "step": 12875
    },
    {
      "epoch": 3.0654645003868355,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 12876
    },
    {
      "epoch": 3.065702553115515,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 12877
    },
    {
      "epoch": 3.0659406058441947,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 12878
    },
    {
      "epoch": 3.066178658572874,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 12879
    },
    {
      "epoch": 3.0664167113015535,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 12880
    },
    {
      "epoch": 3.0666547640302326,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 12881
    },
    {
      "epoch": 3.0668928167589122,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 12882
    },
    {
      "epoch": 3.0671308694875914,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 12883
    },
    {
      "epoch": 3.067368922216271,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 12884
    },
    {
      "epoch": 3.06760697494495,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 12885
    },
    {
      "epoch": 3.0678450276736298,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4314,
      "step": 12886
    },
    {
      "epoch": 3.068083080402309,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 12887
    },
    {
      "epoch": 3.0683211331309885,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 12888
    },
    {
      "epoch": 3.0685591858596677,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 12889
    },
    {
      "epoch": 3.0687972385883473,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 12890
    },
    {
      "epoch": 3.069035291317027,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2842,
      "step": 12891
    },
    {
      "epoch": 3.069273344045706,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 12892
    },
    {
      "epoch": 3.0695113967743857,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 12893
    },
    {
      "epoch": 3.069749449503065,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 12894
    },
    {
      "epoch": 3.0699875022317444,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 12895
    },
    {
      "epoch": 3.0702255549604236,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 12896
    },
    {
      "epoch": 3.070463607689103,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 12897
    },
    {
      "epoch": 3.0707016604177824,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 12898
    },
    {
      "epoch": 3.070939713146462,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 12899
    },
    {
      "epoch": 3.071177765875141,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 12900
    },
    {
      "epoch": 3.071177765875141,
      "eval_runtime": 0.5203,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 12900
    },
    {
      "epoch": 3.0714158186038207,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 12901
    },
    {
      "epoch": 3.0716538713325003,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 12902
    },
    {
      "epoch": 3.0718919240611795,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 12903
    },
    {
      "epoch": 3.072129976789859,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 12904
    },
    {
      "epoch": 3.0723680295185383,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 12905
    },
    {
      "epoch": 3.072606082247218,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 12906
    },
    {
      "epoch": 3.072844134975897,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 12907
    },
    {
      "epoch": 3.0730821877045766,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 12908
    },
    {
      "epoch": 3.073320240433256,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 12909
    },
    {
      "epoch": 3.0735582931619354,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 12910
    },
    {
      "epoch": 3.0737963458906146,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 12911
    },
    {
      "epoch": 3.074034398619294,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3174,
      "step": 12912
    },
    {
      "epoch": 3.074272451347974,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 12913
    },
    {
      "epoch": 3.074510504076653,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 12914
    },
    {
      "epoch": 3.0747485568053325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 12915
    },
    {
      "epoch": 3.0749866095340117,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 12916
    },
    {
      "epoch": 3.0752246622626913,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 12917
    },
    {
      "epoch": 3.0754627149913705,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 12918
    },
    {
      "epoch": 3.07570076772005,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 12919
    },
    {
      "epoch": 3.0759388204487292,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 12920
    },
    {
      "epoch": 3.076176873177409,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 12921
    },
    {
      "epoch": 3.076414925906088,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 12922
    },
    {
      "epoch": 3.0766529786347676,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 12923
    },
    {
      "epoch": 3.076891031363447,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 12924
    },
    {
      "epoch": 3.0771290840921264,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 12925
    },
    {
      "epoch": 3.077367136820806,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 12926
    },
    {
      "epoch": 3.077605189549485,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2103,
      "step": 12927
    },
    {
      "epoch": 3.0778432422781647,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 12928
    },
    {
      "epoch": 3.078081295006844,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 12929
    },
    {
      "epoch": 3.0783193477355235,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 12930
    },
    {
      "epoch": 3.0785574004642027,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 12931
    },
    {
      "epoch": 3.0787954531928823,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.361,
      "step": 12932
    },
    {
      "epoch": 3.0790335059215614,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 12933
    },
    {
      "epoch": 3.079271558650241,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 12934
    },
    {
      "epoch": 3.0795096113789207,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 12935
    },
    {
      "epoch": 3.0797476641076,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 12936
    },
    {
      "epoch": 3.0799857168362794,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 12937
    },
    {
      "epoch": 3.0802237695649586,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3418,
      "step": 12938
    },
    {
      "epoch": 3.080461822293638,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 12939
    },
    {
      "epoch": 3.0806998750223173,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 12940
    },
    {
      "epoch": 3.080937927750997,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 12941
    },
    {
      "epoch": 3.081175980479676,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 12942
    },
    {
      "epoch": 3.0814140332083557,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 12943
    },
    {
      "epoch": 3.081652085937035,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 12944
    },
    {
      "epoch": 3.0818901386657145,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 12945
    },
    {
      "epoch": 3.0821281913943936,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.286,
      "step": 12946
    },
    {
      "epoch": 3.0823662441230733,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 12947
    },
    {
      "epoch": 3.082604296851753,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 12948
    },
    {
      "epoch": 3.082842349580432,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 12949
    },
    {
      "epoch": 3.0830804023091116,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 12950
    },
    {
      "epoch": 3.083318455037791,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 12951
    },
    {
      "epoch": 3.0835565077664704,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 12952
    },
    {
      "epoch": 3.0837945604951496,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 12953
    },
    {
      "epoch": 3.084032613223829,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 12954
    },
    {
      "epoch": 3.0842706659525083,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 12955
    },
    {
      "epoch": 3.084508718681188,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 12956
    },
    {
      "epoch": 3.084746771409867,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 12957
    },
    {
      "epoch": 3.0849848241385467,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 12958
    },
    {
      "epoch": 3.0852228768672263,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 12959
    },
    {
      "epoch": 3.0854609295959055,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 12960
    },
    {
      "epoch": 3.085698982324585,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.5139,
      "step": 12961
    },
    {
      "epoch": 3.085937035053264,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 12962
    },
    {
      "epoch": 3.086175087781944,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 12963
    },
    {
      "epoch": 3.086413140510623,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 12964
    },
    {
      "epoch": 3.0866511932393026,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3903,
      "step": 12965
    },
    {
      "epoch": 3.0868892459679818,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 12966
    },
    {
      "epoch": 3.0871272986966614,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 12967
    },
    {
      "epoch": 3.0873653514253405,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 12968
    },
    {
      "epoch": 3.08760340415402,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 12969
    },
    {
      "epoch": 3.0878414568826997,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 12970
    },
    {
      "epoch": 3.088079509611379,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 12971
    },
    {
      "epoch": 3.0883175623400585,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 12972
    },
    {
      "epoch": 3.0885556150687377,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 12973
    },
    {
      "epoch": 3.0887936677974173,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 12974
    },
    {
      "epoch": 3.0890317205260964,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 12975
    },
    {
      "epoch": 3.089269773254776,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 12976
    },
    {
      "epoch": 3.089507825983455,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 12977
    },
    {
      "epoch": 3.089745878712135,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.5466,
      "step": 12978
    },
    {
      "epoch": 3.089983931440814,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 12979
    },
    {
      "epoch": 3.0902219841694936,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 12980
    },
    {
      "epoch": 3.0904600368981727,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 12981
    },
    {
      "epoch": 3.0906980896268523,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 12982
    },
    {
      "epoch": 3.090936142355532,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 12983
    },
    {
      "epoch": 3.091174195084211,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 12984
    },
    {
      "epoch": 3.0914122478128907,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 12985
    },
    {
      "epoch": 3.09165030054157,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 12986
    },
    {
      "epoch": 3.0918883532702495,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 12987
    },
    {
      "epoch": 3.0921264059989286,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 12988
    },
    {
      "epoch": 3.0923644587276082,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 12989
    },
    {
      "epoch": 3.0926025114562874,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 12990
    },
    {
      "epoch": 3.092840564184967,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 12991
    },
    {
      "epoch": 3.093078616913646,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 12992
    },
    {
      "epoch": 3.0933166696423258,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 12993
    },
    {
      "epoch": 3.0935547223710054,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 12994
    },
    {
      "epoch": 3.0937927750996845,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2474,
      "step": 12995
    },
    {
      "epoch": 3.094030827828364,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3288,
      "step": 12996
    },
    {
      "epoch": 3.0942688805570433,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 12997
    },
    {
      "epoch": 3.094506933285723,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 12998
    },
    {
      "epoch": 3.094744986014402,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 12999
    },
    {
      "epoch": 3.0949830387430817,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.4632,
      "step": 13000
    },
    {
      "epoch": 3.0949830387430817,
      "eval_runtime": 0.5161,
      "eval_samples_per_second": 1.937,
      "eval_steps_per_second": 1.937,
      "step": 13000
    },
    {
      "epoch": 3.095221091471761,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 13001
    },
    {
      "epoch": 3.0954591442004404,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 13002
    },
    {
      "epoch": 3.0956971969291196,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 13003
    },
    {
      "epoch": 3.095935249657799,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 13004
    },
    {
      "epoch": 3.096173302386479,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 13005
    },
    {
      "epoch": 3.096411355115158,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 13006
    },
    {
      "epoch": 3.0966494078438376,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3921,
      "step": 13007
    },
    {
      "epoch": 3.0968874605725167,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 13008
    },
    {
      "epoch": 3.0971255133011963,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 13009
    },
    {
      "epoch": 3.0973635660298755,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2561,
      "step": 13010
    },
    {
      "epoch": 3.097601618758555,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 13011
    },
    {
      "epoch": 3.0978396714872343,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 13012
    },
    {
      "epoch": 3.098077724215914,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 13013
    },
    {
      "epoch": 3.098315776944593,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 13014
    },
    {
      "epoch": 3.0985538296732726,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7365,
      "step": 13015
    },
    {
      "epoch": 3.0987918824019522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9751,
      "step": 13016
    },
    {
      "epoch": 3.0990299351306314,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3473,
      "step": 13017
    },
    {
      "epoch": 3.099267987859311,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 13018
    },
    {
      "epoch": 3.09950604058799,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 13019
    },
    {
      "epoch": 3.09974409331667,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 13020
    },
    {
      "epoch": 3.099982146045349,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3778,
      "step": 13021
    },
    {
      "epoch": 3.1002201987740285,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 13022
    },
    {
      "epoch": 3.1004582515027077,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 13023
    },
    {
      "epoch": 3.1006963042313873,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 13024
    },
    {
      "epoch": 3.1009343569600665,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 13025
    },
    {
      "epoch": 3.101172409688746,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 13026
    },
    {
      "epoch": 3.1014104624174257,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 13027
    },
    {
      "epoch": 3.101648515146105,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 13028
    },
    {
      "epoch": 3.1018865678747845,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 13029
    },
    {
      "epoch": 3.1021246206034636,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 13030
    },
    {
      "epoch": 3.102362673332143,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 13031
    },
    {
      "epoch": 3.1026007260608224,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 13032
    },
    {
      "epoch": 3.102838778789502,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 13033
    },
    {
      "epoch": 3.103076831518181,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 13034
    },
    {
      "epoch": 3.1033148842468607,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 13035
    },
    {
      "epoch": 3.10355293697554,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6717,
      "step": 13036
    },
    {
      "epoch": 3.1037909897042195,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3959,
      "step": 13037
    },
    {
      "epoch": 3.1040290424328987,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 13038
    },
    {
      "epoch": 3.1042670951615783,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 13039
    },
    {
      "epoch": 3.104505147890258,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 13040
    },
    {
      "epoch": 3.104743200618937,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 13041
    },
    {
      "epoch": 3.1049812533476167,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3167,
      "step": 13042
    },
    {
      "epoch": 3.105219306076296,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 13043
    },
    {
      "epoch": 3.1054573588049754,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 13044
    },
    {
      "epoch": 3.1056954115336546,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 13045
    },
    {
      "epoch": 3.105933464262334,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 13046
    },
    {
      "epoch": 3.1061715169910133,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4795,
      "step": 13047
    },
    {
      "epoch": 3.106409569719693,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 13048
    },
    {
      "epoch": 3.106647622448372,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 13049
    },
    {
      "epoch": 3.1068856751770517,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 13050
    },
    {
      "epoch": 3.1071237279057313,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 13051
    },
    {
      "epoch": 3.1073617806344105,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2188,
      "step": 13052
    },
    {
      "epoch": 3.10759983336309,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 13053
    },
    {
      "epoch": 3.1078378860917693,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3639,
      "step": 13054
    },
    {
      "epoch": 3.108075938820449,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 13055
    },
    {
      "epoch": 3.108313991549128,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 13056
    },
    {
      "epoch": 3.1085520442778076,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 13057
    },
    {
      "epoch": 3.108790097006487,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 13058
    },
    {
      "epoch": 3.1090281497351664,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 13059
    },
    {
      "epoch": 3.1092662024638456,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 13060
    },
    {
      "epoch": 3.109504255192525,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 13061
    },
    {
      "epoch": 3.1097423079212048,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4842,
      "step": 13062
    },
    {
      "epoch": 3.109980360649884,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 13063
    },
    {
      "epoch": 3.1102184133785635,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 13064
    },
    {
      "epoch": 3.1104564661072427,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 13065
    },
    {
      "epoch": 3.1106945188359223,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 13066
    },
    {
      "epoch": 3.1109325715646015,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 13067
    },
    {
      "epoch": 3.111170624293281,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 13068
    },
    {
      "epoch": 3.1114086770219602,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 13069
    },
    {
      "epoch": 3.11164672975064,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 13070
    },
    {
      "epoch": 3.111884782479319,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 13071
    },
    {
      "epoch": 3.1121228352079986,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 13072
    },
    {
      "epoch": 3.1123608879366778,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 13073
    },
    {
      "epoch": 3.1125989406653574,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 13074
    },
    {
      "epoch": 3.112836993394037,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3487,
      "step": 13075
    },
    {
      "epoch": 3.113075046122716,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.471,
      "step": 13076
    },
    {
      "epoch": 3.1133130988513957,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 13077
    },
    {
      "epoch": 3.113551151580075,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 13078
    },
    {
      "epoch": 3.1137892043087545,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 13079
    },
    {
      "epoch": 3.1140272570374337,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 13080
    },
    {
      "epoch": 3.1142653097661133,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 13081
    },
    {
      "epoch": 3.1145033624947924,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2455,
      "step": 13082
    },
    {
      "epoch": 3.114741415223472,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 13083
    },
    {
      "epoch": 3.114979467952151,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4885,
      "step": 13084
    },
    {
      "epoch": 3.115217520680831,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 13085
    },
    {
      "epoch": 3.1154555734095104,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3845,
      "step": 13086
    },
    {
      "epoch": 3.1156936261381896,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 13087
    },
    {
      "epoch": 3.115931678866869,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 13088
    },
    {
      "epoch": 3.1161697315955483,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 13089
    },
    {
      "epoch": 3.116407784324228,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 13090
    },
    {
      "epoch": 3.116645837052907,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 13091
    },
    {
      "epoch": 3.1168838897815867,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 13092
    },
    {
      "epoch": 3.117121942510266,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 13093
    },
    {
      "epoch": 3.1173599952389455,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 13094
    },
    {
      "epoch": 3.1175980479676246,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 13095
    },
    {
      "epoch": 3.1178361006963042,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 13096
    },
    {
      "epoch": 3.118074153424984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 13097
    },
    {
      "epoch": 3.118312206153663,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 13098
    },
    {
      "epoch": 3.1185502588823426,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 13099
    },
    {
      "epoch": 3.1187883116110218,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 13100
    },
    {
      "epoch": 3.1187883116110218,
      "eval_runtime": 0.5242,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 13100
    },
    {
      "epoch": 3.1190263643397014,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 13101
    },
    {
      "epoch": 3.1192644170683805,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 13102
    },
    {
      "epoch": 3.11950246979706,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 13103
    },
    {
      "epoch": 3.1197405225257393,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 13104
    },
    {
      "epoch": 3.119978575254419,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 13105
    },
    {
      "epoch": 3.120216627983098,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 13106
    },
    {
      "epoch": 3.1204546807117777,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 13107
    },
    {
      "epoch": 3.120692733440457,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 13108
    },
    {
      "epoch": 3.1209307861691364,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 13109
    },
    {
      "epoch": 3.121168838897816,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3276,
      "step": 13110
    },
    {
      "epoch": 3.121406891626495,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 13111
    },
    {
      "epoch": 3.121644944355175,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 13112
    },
    {
      "epoch": 3.121882997083854,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 13113
    },
    {
      "epoch": 3.1221210498125336,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 13114
    },
    {
      "epoch": 3.1223591025412127,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 13115
    },
    {
      "epoch": 3.1225971552698923,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 13116
    },
    {
      "epoch": 3.1228352079985715,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2814,
      "step": 13117
    },
    {
      "epoch": 3.123073260727251,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 13118
    },
    {
      "epoch": 3.1233113134559307,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 13119
    },
    {
      "epoch": 3.12354936618461,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 13120
    },
    {
      "epoch": 3.1237874189132895,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.5536,
      "step": 13121
    },
    {
      "epoch": 3.1240254716419686,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 13122
    },
    {
      "epoch": 3.1242635243706482,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 13123
    },
    {
      "epoch": 3.1245015770993274,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 13124
    },
    {
      "epoch": 3.124739629828007,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 13125
    },
    {
      "epoch": 3.124977682556686,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 13126
    },
    {
      "epoch": 3.125215735285366,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 13127
    },
    {
      "epoch": 3.125453788014045,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 13128
    },
    {
      "epoch": 3.1256918407427245,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 13129
    },
    {
      "epoch": 3.1259298934714037,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 13130
    },
    {
      "epoch": 3.1261679462000833,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 13131
    },
    {
      "epoch": 3.126405998928763,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 13132
    },
    {
      "epoch": 3.126644051657442,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 13133
    },
    {
      "epoch": 3.1268821043861217,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4407,
      "step": 13134
    },
    {
      "epoch": 3.127120157114801,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 13135
    },
    {
      "epoch": 3.1273582098434805,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 13136
    },
    {
      "epoch": 3.1275962625721596,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 13137
    },
    {
      "epoch": 3.127834315300839,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 13138
    },
    {
      "epoch": 3.1280723680295184,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 13139
    },
    {
      "epoch": 3.128310420758198,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 13140
    },
    {
      "epoch": 3.128548473486877,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4392,
      "step": 13141
    },
    {
      "epoch": 3.1287865262155568,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 13142
    },
    {
      "epoch": 3.129024578944236,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 13143
    },
    {
      "epoch": 3.1292626316729155,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 13144
    },
    {
      "epoch": 3.129500684401595,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 13145
    },
    {
      "epoch": 3.1297387371302743,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 13146
    },
    {
      "epoch": 3.129976789858954,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 13147
    },
    {
      "epoch": 3.130214842587633,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 13148
    },
    {
      "epoch": 3.1304528953163127,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 13149
    },
    {
      "epoch": 3.130690948044992,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 13150
    },
    {
      "epoch": 3.1309290007736714,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 13151
    },
    {
      "epoch": 3.1311670535023506,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 13152
    },
    {
      "epoch": 3.13140510623103,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 13153
    },
    {
      "epoch": 3.13164315895971,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 13154
    },
    {
      "epoch": 3.131881211688389,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 13155
    },
    {
      "epoch": 3.1321192644170686,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4193,
      "step": 13156
    },
    {
      "epoch": 3.1323573171457477,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 13157
    },
    {
      "epoch": 3.1325953698744273,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 13158
    },
    {
      "epoch": 3.1328334226031065,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 13159
    },
    {
      "epoch": 3.133071475331786,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 13160
    },
    {
      "epoch": 3.1333095280604653,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 13161
    },
    {
      "epoch": 3.133547580789145,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 13162
    },
    {
      "epoch": 3.133785633517824,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 13163
    },
    {
      "epoch": 3.1340236862465036,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 13164
    },
    {
      "epoch": 3.134261738975183,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 13165
    },
    {
      "epoch": 3.1344997917038624,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 13166
    },
    {
      "epoch": 3.134737844432542,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 13167
    },
    {
      "epoch": 3.134975897161221,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 13168
    },
    {
      "epoch": 3.1352139498899008,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 13169
    },
    {
      "epoch": 3.13545200261858,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 13170
    },
    {
      "epoch": 3.1356900553472595,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.5474,
      "step": 13171
    },
    {
      "epoch": 3.1359281080759387,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 13172
    },
    {
      "epoch": 3.1361661608046183,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 13173
    },
    {
      "epoch": 3.1364042135332975,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 13174
    },
    {
      "epoch": 3.136642266261977,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 13175
    },
    {
      "epoch": 3.1368803189906562,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 13176
    },
    {
      "epoch": 3.137118371719336,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3811,
      "step": 13177
    },
    {
      "epoch": 3.1373564244480154,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 13178
    },
    {
      "epoch": 3.1375944771766946,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 13179
    },
    {
      "epoch": 3.137832529905374,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 13180
    },
    {
      "epoch": 3.1380705826340534,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 13181
    },
    {
      "epoch": 3.138308635362733,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 13182
    },
    {
      "epoch": 3.138546688091412,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 13183
    },
    {
      "epoch": 3.1387847408200917,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5456,
      "step": 13184
    },
    {
      "epoch": 3.139022793548771,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 13185
    },
    {
      "epoch": 3.1392608462774505,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3488,
      "step": 13186
    },
    {
      "epoch": 3.1394988990061297,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 13187
    },
    {
      "epoch": 3.1397369517348093,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 13188
    },
    {
      "epoch": 3.139975004463489,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 13189
    },
    {
      "epoch": 3.140213057192168,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 13190
    },
    {
      "epoch": 3.1404511099208476,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 13191
    },
    {
      "epoch": 3.140689162649527,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 13192
    },
    {
      "epoch": 3.1409272153782064,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 13193
    },
    {
      "epoch": 3.1411652681068856,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 13194
    },
    {
      "epoch": 3.141403320835565,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4096,
      "step": 13195
    },
    {
      "epoch": 3.1416413735642443,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3579,
      "step": 13196
    },
    {
      "epoch": 3.141879426292924,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 13197
    },
    {
      "epoch": 3.142117479021603,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 13198
    },
    {
      "epoch": 3.1423555317502827,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 13199
    },
    {
      "epoch": 3.142593584478962,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 13200
    },
    {
      "epoch": 3.142593584478962,
      "eval_runtime": 0.5099,
      "eval_samples_per_second": 1.961,
      "eval_steps_per_second": 1.961,
      "step": 13200
    },
    {
      "epoch": 3.1428316372076415,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 13201
    },
    {
      "epoch": 3.143069689936321,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 13202
    },
    {
      "epoch": 3.1433077426650002,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 13203
    },
    {
      "epoch": 3.14354579539368,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 13204
    },
    {
      "epoch": 3.143783848122359,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 13205
    },
    {
      "epoch": 3.1440219008510386,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 13206
    },
    {
      "epoch": 3.1442599535797178,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 13207
    },
    {
      "epoch": 3.1444980063083974,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 13208
    },
    {
      "epoch": 3.1447360590370765,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 13209
    },
    {
      "epoch": 3.144974111765756,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 13210
    },
    {
      "epoch": 3.1452121644944357,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 13211
    },
    {
      "epoch": 3.145450217223115,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 13212
    },
    {
      "epoch": 3.1456882699517945,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 13213
    },
    {
      "epoch": 3.1459263226804737,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 13214
    },
    {
      "epoch": 3.1461643754091533,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 13215
    },
    {
      "epoch": 3.1464024281378324,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 13216
    },
    {
      "epoch": 3.146640480866512,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4352,
      "step": 13217
    },
    {
      "epoch": 3.146878533595191,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 13218
    },
    {
      "epoch": 3.147116586323871,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 13219
    },
    {
      "epoch": 3.14735463905255,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 13220
    },
    {
      "epoch": 3.1475926917812296,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 13221
    },
    {
      "epoch": 3.1478307445099087,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 13222
    },
    {
      "epoch": 3.1480687972385883,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 13223
    },
    {
      "epoch": 3.148306849967268,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 13224
    },
    {
      "epoch": 3.148544902695947,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 13225
    },
    {
      "epoch": 3.1487829554246267,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 13226
    },
    {
      "epoch": 3.149021008153306,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 13227
    },
    {
      "epoch": 3.1492590608819855,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 13228
    },
    {
      "epoch": 3.1494971136106646,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 13229
    },
    {
      "epoch": 3.1497351663393443,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 13230
    },
    {
      "epoch": 3.1499732190680234,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 13231
    },
    {
      "epoch": 3.150211271796703,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 13232
    },
    {
      "epoch": 3.150449324525382,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 13233
    },
    {
      "epoch": 3.150687377254062,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 13234
    },
    {
      "epoch": 3.150925429982741,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.319,
      "step": 13235
    },
    {
      "epoch": 3.1511634827114205,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 13236
    },
    {
      "epoch": 3.1514015354401,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 13237
    },
    {
      "epoch": 3.1516395881687793,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 13238
    },
    {
      "epoch": 3.151877640897459,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 13239
    },
    {
      "epoch": 3.152115693626138,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3117,
      "step": 13240
    },
    {
      "epoch": 3.1523537463548177,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 13241
    },
    {
      "epoch": 3.152591799083497,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 13242
    },
    {
      "epoch": 3.1528298518121765,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 13243
    },
    {
      "epoch": 3.1530679045408556,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 13244
    },
    {
      "epoch": 3.153305957269535,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 13245
    },
    {
      "epoch": 3.153544009998215,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 13246
    },
    {
      "epoch": 3.153782062726894,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 13247
    },
    {
      "epoch": 3.1540201154555736,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 13248
    },
    {
      "epoch": 3.1542581681842528,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 13249
    },
    {
      "epoch": 3.1544962209129324,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 13250
    },
    {
      "epoch": 3.1547342736416115,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 13251
    },
    {
      "epoch": 3.154972326370291,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 13252
    },
    {
      "epoch": 3.1552103790989703,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 13253
    },
    {
      "epoch": 3.15544843182765,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 13254
    },
    {
      "epoch": 3.155686484556329,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 13255
    },
    {
      "epoch": 3.1559245372850087,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 13256
    },
    {
      "epoch": 3.156162590013688,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 13257
    },
    {
      "epoch": 3.1564006427423674,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 13258
    },
    {
      "epoch": 3.156638695471047,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 13259
    },
    {
      "epoch": 3.156876748199726,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 13260
    },
    {
      "epoch": 3.157114800928406,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 13261
    },
    {
      "epoch": 3.157352853657085,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 13262
    },
    {
      "epoch": 3.1575909063857646,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 13263
    },
    {
      "epoch": 3.1578289591144437,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 13264
    },
    {
      "epoch": 3.1580670118431233,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 13265
    },
    {
      "epoch": 3.1583050645718025,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 13266
    },
    {
      "epoch": 3.158543117300482,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 13267
    },
    {
      "epoch": 3.1587811700291613,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 13268
    },
    {
      "epoch": 3.159019222757841,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3847,
      "step": 13269
    },
    {
      "epoch": 3.1592572754865205,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 13270
    },
    {
      "epoch": 3.1594953282151996,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 13271
    },
    {
      "epoch": 3.1597333809438792,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 13272
    },
    {
      "epoch": 3.1599714336725584,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 13273
    },
    {
      "epoch": 3.160209486401238,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 13274
    },
    {
      "epoch": 3.160447539129917,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 13275
    },
    {
      "epoch": 3.1606855918585968,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 13276
    },
    {
      "epoch": 3.160923644587276,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 13277
    },
    {
      "epoch": 3.1611616973159555,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 13278
    },
    {
      "epoch": 3.1613997500446347,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 13279
    },
    {
      "epoch": 3.1616378027733143,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 13280
    },
    {
      "epoch": 3.161875855501994,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 13281
    },
    {
      "epoch": 3.162113908230673,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 13282
    },
    {
      "epoch": 3.1623519609593527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 13283
    },
    {
      "epoch": 3.162590013688032,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.342,
      "step": 13284
    },
    {
      "epoch": 3.1628280664167114,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 13285
    },
    {
      "epoch": 3.1630661191453906,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4613,
      "step": 13286
    },
    {
      "epoch": 3.16330417187407,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 13287
    },
    {
      "epoch": 3.1635422246027494,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3504,
      "step": 13288
    },
    {
      "epoch": 3.163780277331429,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 13289
    },
    {
      "epoch": 3.164018330060108,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 13290
    },
    {
      "epoch": 3.1642563827887877,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 13291
    },
    {
      "epoch": 3.164494435517467,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 13292
    },
    {
      "epoch": 3.1647324882461465,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 13293
    },
    {
      "epoch": 3.164970540974826,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3761,
      "step": 13294
    },
    {
      "epoch": 3.1652085937035053,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 13295
    },
    {
      "epoch": 3.165446646432185,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 13296
    },
    {
      "epoch": 3.165684699160864,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 13297
    },
    {
      "epoch": 3.1659227518895436,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 13298
    },
    {
      "epoch": 3.166160804618223,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 13299
    },
    {
      "epoch": 3.1663988573469024,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 13300
    },
    {
      "epoch": 3.1663988573469024,
      "eval_runtime": 0.5112,
      "eval_samples_per_second": 1.956,
      "eval_steps_per_second": 1.956,
      "step": 13300
    },
    {
      "epoch": 3.1666369100755816,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 13301
    },
    {
      "epoch": 3.166874962804261,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 13302
    },
    {
      "epoch": 3.167113015532941,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 13303
    },
    {
      "epoch": 3.16735106826162,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 13304
    },
    {
      "epoch": 3.1675891209902995,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 13305
    },
    {
      "epoch": 3.1678271737189787,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3496,
      "step": 13306
    },
    {
      "epoch": 3.1680652264476583,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 13307
    },
    {
      "epoch": 3.1683032791763375,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 13308
    },
    {
      "epoch": 3.168541331905017,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 13309
    },
    {
      "epoch": 3.1687793846336962,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3564,
      "step": 13310
    },
    {
      "epoch": 3.169017437362376,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2773,
      "step": 13311
    },
    {
      "epoch": 3.169255490091055,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 13312
    },
    {
      "epoch": 3.1694935428197346,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 13313
    },
    {
      "epoch": 3.1697315955484138,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 13314
    },
    {
      "epoch": 3.1699696482770934,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 13315
    },
    {
      "epoch": 3.170207701005773,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4607,
      "step": 13316
    },
    {
      "epoch": 3.170445753734452,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 13317
    },
    {
      "epoch": 3.1706838064631317,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 13318
    },
    {
      "epoch": 3.170921859191811,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.5203,
      "step": 13319
    },
    {
      "epoch": 3.1711599119204905,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 13320
    },
    {
      "epoch": 3.1713979646491697,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 13321
    },
    {
      "epoch": 3.1716360173778493,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 13322
    },
    {
      "epoch": 3.1718740701065284,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 13323
    },
    {
      "epoch": 3.172112122835208,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 13324
    },
    {
      "epoch": 3.172350175563887,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 13325
    },
    {
      "epoch": 3.172588228292567,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3164,
      "step": 13326
    },
    {
      "epoch": 3.172826281021246,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3998,
      "step": 13327
    },
    {
      "epoch": 3.1730643337499256,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 13328
    },
    {
      "epoch": 3.173302386478605,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 13329
    },
    {
      "epoch": 3.1735404392072843,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 13330
    },
    {
      "epoch": 3.173778491935964,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 13331
    },
    {
      "epoch": 3.174016544664643,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 13332
    },
    {
      "epoch": 3.1742545973933227,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 13333
    },
    {
      "epoch": 3.174492650122002,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 13334
    },
    {
      "epoch": 3.1747307028506815,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 13335
    },
    {
      "epoch": 3.1749687555793606,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 13336
    },
    {
      "epoch": 3.1752068083080403,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 13337
    },
    {
      "epoch": 3.17544486103672,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 13338
    },
    {
      "epoch": 3.175682913765399,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 13339
    },
    {
      "epoch": 3.1759209664940786,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 13340
    },
    {
      "epoch": 3.176159019222758,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.4699,
      "step": 13341
    },
    {
      "epoch": 3.1763970719514374,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 13342
    },
    {
      "epoch": 3.1766351246801166,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 13343
    },
    {
      "epoch": 3.176873177408796,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 13344
    },
    {
      "epoch": 3.1771112301374753,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 13345
    },
    {
      "epoch": 3.177349282866155,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 13346
    },
    {
      "epoch": 3.177587335594834,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 13347
    },
    {
      "epoch": 3.1778253883235137,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 13348
    },
    {
      "epoch": 3.178063441052193,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 13349
    },
    {
      "epoch": 3.1783014937808725,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 13350
    },
    {
      "epoch": 3.178539546509552,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 13351
    },
    {
      "epoch": 3.178777599238231,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 13352
    },
    {
      "epoch": 3.179015651966911,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 13353
    },
    {
      "epoch": 3.17925370469559,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 13354
    },
    {
      "epoch": 3.1794917574242696,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 13355
    },
    {
      "epoch": 3.1797298101529488,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 13356
    },
    {
      "epoch": 3.1799678628816284,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 13357
    },
    {
      "epoch": 3.1802059156103075,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 13358
    },
    {
      "epoch": 3.180443968338987,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 13359
    },
    {
      "epoch": 3.1806820210676663,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 13360
    },
    {
      "epoch": 3.180920073796346,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 13361
    },
    {
      "epoch": 3.1811581265250255,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 13362
    },
    {
      "epoch": 3.1813961792537047,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2103,
      "step": 13363
    },
    {
      "epoch": 3.1816342319823843,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 13364
    },
    {
      "epoch": 3.1818722847110634,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 13365
    },
    {
      "epoch": 3.182110337439743,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 13366
    },
    {
      "epoch": 3.182348390168422,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 13367
    },
    {
      "epoch": 3.182586442897102,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 13368
    },
    {
      "epoch": 3.182824495625781,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 13369
    },
    {
      "epoch": 3.1830625483544606,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 13370
    },
    {
      "epoch": 3.1833006010831397,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 13371
    },
    {
      "epoch": 3.1835386538118193,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 13372
    },
    {
      "epoch": 3.183776706540499,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 13373
    },
    {
      "epoch": 3.184014759269178,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 13374
    },
    {
      "epoch": 3.1842528119978577,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 13375
    },
    {
      "epoch": 3.184490864726537,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 13376
    },
    {
      "epoch": 3.1847289174552165,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 13377
    },
    {
      "epoch": 3.1849669701838956,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 13378
    },
    {
      "epoch": 3.1852050229125752,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 13379
    },
    {
      "epoch": 3.1854430756412544,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 13380
    },
    {
      "epoch": 3.185681128369934,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 13381
    },
    {
      "epoch": 3.185919181098613,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 13382
    },
    {
      "epoch": 3.1861572338272928,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 13383
    },
    {
      "epoch": 3.186395286555972,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 13384
    },
    {
      "epoch": 3.1866333392846515,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4543,
      "step": 13385
    },
    {
      "epoch": 3.186871392013331,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 13386
    },
    {
      "epoch": 3.1871094447420103,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 13387
    },
    {
      "epoch": 3.18734749747069,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 13388
    },
    {
      "epoch": 3.187585550199369,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3458,
      "step": 13389
    },
    {
      "epoch": 3.1878236029280487,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 13390
    },
    {
      "epoch": 3.188061655656728,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 13391
    },
    {
      "epoch": 3.1882997083854074,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 13392
    },
    {
      "epoch": 3.1885377611140866,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 13393
    },
    {
      "epoch": 3.188775813842766,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 13394
    },
    {
      "epoch": 3.189013866571446,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 13395
    },
    {
      "epoch": 3.189251919300125,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 13396
    },
    {
      "epoch": 3.1894899720288046,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 13397
    },
    {
      "epoch": 3.1897280247574837,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3242,
      "step": 13398
    },
    {
      "epoch": 3.1899660774861633,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 13399
    },
    {
      "epoch": 3.1902041302148425,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 13400
    },
    {
      "epoch": 3.1902041302148425,
      "eval_runtime": 0.5067,
      "eval_samples_per_second": 1.973,
      "eval_steps_per_second": 1.973,
      "step": 13400
    },
    {
      "epoch": 3.190442182943522,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 13401
    },
    {
      "epoch": 3.1906802356722013,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 13402
    },
    {
      "epoch": 3.190918288400881,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 13403
    },
    {
      "epoch": 3.19115634112956,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 13404
    },
    {
      "epoch": 3.1913943938582396,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 13405
    },
    {
      "epoch": 3.191632446586919,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 13406
    },
    {
      "epoch": 3.1918704993155984,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3361,
      "step": 13407
    },
    {
      "epoch": 3.192108552044278,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 13408
    },
    {
      "epoch": 3.192346604772957,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 13409
    },
    {
      "epoch": 3.192584657501637,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4477,
      "step": 13410
    },
    {
      "epoch": 3.192822710230316,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4179,
      "step": 13411
    },
    {
      "epoch": 3.1930607629589955,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 13412
    },
    {
      "epoch": 3.1932988156876747,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 13413
    },
    {
      "epoch": 3.1935368684163543,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 13414
    },
    {
      "epoch": 3.1937749211450335,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 13415
    },
    {
      "epoch": 3.194012973873713,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 13416
    },
    {
      "epoch": 3.1942510266023922,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 13417
    },
    {
      "epoch": 3.194489079331072,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2752,
      "step": 13418
    },
    {
      "epoch": 3.194727132059751,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 13419
    },
    {
      "epoch": 3.1949651847884306,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 13420
    },
    {
      "epoch": 3.19520323751711,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 13421
    },
    {
      "epoch": 3.1954412902457894,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7431,
      "step": 13422
    },
    {
      "epoch": 3.195679342974469,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 13423
    },
    {
      "epoch": 3.195917395703148,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 13424
    },
    {
      "epoch": 3.1961554484318278,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 13425
    },
    {
      "epoch": 3.196393501160507,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 13426
    },
    {
      "epoch": 3.1966315538891865,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4193,
      "step": 13427
    },
    {
      "epoch": 3.1968696066178657,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 13428
    },
    {
      "epoch": 3.1971076593465453,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 13429
    },
    {
      "epoch": 3.197345712075225,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 13430
    },
    {
      "epoch": 3.197583764803904,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 13431
    },
    {
      "epoch": 3.1978218175325837,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 13432
    },
    {
      "epoch": 3.198059870261263,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 13433
    },
    {
      "epoch": 3.1982979229899424,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 13434
    },
    {
      "epoch": 3.1985359757186216,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 13435
    },
    {
      "epoch": 3.198774028447301,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 13436
    },
    {
      "epoch": 3.1990120811759803,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 13437
    },
    {
      "epoch": 3.19925013390466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 13438
    },
    {
      "epoch": 3.199488186633339,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 13439
    },
    {
      "epoch": 3.1997262393620187,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 13440
    },
    {
      "epoch": 3.199964292090698,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 13441
    },
    {
      "epoch": 3.2002023448193775,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 13442
    },
    {
      "epoch": 3.200440397548057,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 13443
    },
    {
      "epoch": 3.2006784502767363,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 13444
    },
    {
      "epoch": 3.200916503005416,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 13445
    },
    {
      "epoch": 3.201154555734095,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 13446
    },
    {
      "epoch": 3.2013926084627746,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 13447
    },
    {
      "epoch": 3.201630661191454,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 13448
    },
    {
      "epoch": 3.2018687139201334,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 13449
    },
    {
      "epoch": 3.2021067666488126,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 13450
    },
    {
      "epoch": 3.202344819377492,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 13451
    },
    {
      "epoch": 3.2025828721061713,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.6582,
      "step": 13452
    },
    {
      "epoch": 3.202820924834851,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3741,
      "step": 13453
    },
    {
      "epoch": 3.2030589775635305,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 13454
    },
    {
      "epoch": 3.2032970302922097,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 13455
    },
    {
      "epoch": 3.2035350830208893,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 13456
    },
    {
      "epoch": 3.2037731357495685,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 13457
    },
    {
      "epoch": 3.204011188478248,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 13458
    },
    {
      "epoch": 3.2042492412069272,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 13459
    },
    {
      "epoch": 3.204487293935607,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 13460
    },
    {
      "epoch": 3.204725346664286,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 13461
    },
    {
      "epoch": 3.2049633993929656,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 13462
    },
    {
      "epoch": 3.2052014521216448,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 13463
    },
    {
      "epoch": 3.2054395048503244,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 13464
    },
    {
      "epoch": 3.205677557579004,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3984,
      "step": 13465
    },
    {
      "epoch": 3.205915610307683,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3294,
      "step": 13466
    },
    {
      "epoch": 3.2061536630363627,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 13467
    },
    {
      "epoch": 3.206391715765042,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 13468
    },
    {
      "epoch": 3.2066297684937215,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 13469
    },
    {
      "epoch": 3.2068678212224007,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 13470
    },
    {
      "epoch": 3.2071058739510803,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 13471
    },
    {
      "epoch": 3.2073439266797594,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 13472
    },
    {
      "epoch": 3.207581979408439,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 13473
    },
    {
      "epoch": 3.207820032137118,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 13474
    },
    {
      "epoch": 3.208058084865798,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 13475
    },
    {
      "epoch": 3.208296137594477,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 13476
    },
    {
      "epoch": 3.2085341903231566,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 13477
    },
    {
      "epoch": 3.208772243051836,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 13478
    },
    {
      "epoch": 3.2090102957805153,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2923,
      "step": 13479
    },
    {
      "epoch": 3.209248348509195,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 13480
    },
    {
      "epoch": 3.209486401237874,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 13481
    },
    {
      "epoch": 3.2097244539665537,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 13482
    },
    {
      "epoch": 3.209962506695233,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 13483
    },
    {
      "epoch": 3.2102005594239125,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 13484
    },
    {
      "epoch": 3.2104386121525916,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 13485
    },
    {
      "epoch": 3.2106766648812712,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.5826,
      "step": 13486
    },
    {
      "epoch": 3.2109147176099504,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2916,
      "step": 13487
    },
    {
      "epoch": 3.21115277033863,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5893,
      "step": 13488
    },
    {
      "epoch": 3.2113908230673096,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2495,
      "step": 13489
    },
    {
      "epoch": 3.2116288757959888,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 13490
    },
    {
      "epoch": 3.2118669285246684,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 13491
    },
    {
      "epoch": 3.2121049812533475,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 13492
    },
    {
      "epoch": 3.212343033982027,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 13493
    },
    {
      "epoch": 3.2125810867107063,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 13494
    },
    {
      "epoch": 3.212819139439386,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 13495
    },
    {
      "epoch": 3.213057192168065,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 13496
    },
    {
      "epoch": 3.2132952448967447,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 13497
    },
    {
      "epoch": 3.213533297625424,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.6769,
      "step": 13498
    },
    {
      "epoch": 3.2137713503541034,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3217,
      "step": 13499
    },
    {
      "epoch": 3.214009403082783,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 13500
    },
    {
      "epoch": 3.214009403082783,
      "eval_runtime": 0.5263,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 13500
    },
    {
      "epoch": 3.214247455811462,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 13501
    },
    {
      "epoch": 3.214485508540142,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 13502
    },
    {
      "epoch": 3.214723561268821,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 13503
    },
    {
      "epoch": 3.2149616139975006,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 13504
    },
    {
      "epoch": 3.2151996667261797,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 13505
    },
    {
      "epoch": 3.2154377194548593,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 13506
    },
    {
      "epoch": 3.2156757721835385,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 13507
    },
    {
      "epoch": 3.215913824912218,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 13508
    },
    {
      "epoch": 3.2161518776408973,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 13509
    },
    {
      "epoch": 3.216389930369577,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 13510
    },
    {
      "epoch": 3.216627983098256,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 13511
    },
    {
      "epoch": 3.2168660358269356,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 13512
    },
    {
      "epoch": 3.2171040885556152,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 13513
    },
    {
      "epoch": 3.2173421412842944,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 13514
    },
    {
      "epoch": 3.217580194012974,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 13515
    },
    {
      "epoch": 3.217818246741653,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 13516
    },
    {
      "epoch": 3.218056299470333,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 13517
    },
    {
      "epoch": 3.218294352199012,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 13518
    },
    {
      "epoch": 3.2185324049276915,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3331,
      "step": 13519
    },
    {
      "epoch": 3.2187704576563707,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 13520
    },
    {
      "epoch": 3.2190085103850503,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 13521
    },
    {
      "epoch": 3.21924656311373,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 13522
    },
    {
      "epoch": 3.219484615842409,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 13523
    },
    {
      "epoch": 3.2197226685710887,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 13524
    },
    {
      "epoch": 3.219960721299768,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 13525
    },
    {
      "epoch": 3.2201987740284475,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 13526
    },
    {
      "epoch": 3.2204368267571266,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 13527
    },
    {
      "epoch": 3.220674879485806,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 13528
    },
    {
      "epoch": 3.2209129322144854,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 13529
    },
    {
      "epoch": 3.221150984943165,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.5071,
      "step": 13530
    },
    {
      "epoch": 3.221389037671844,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4166,
      "step": 13531
    },
    {
      "epoch": 3.2216270904005238,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 13532
    },
    {
      "epoch": 3.221865143129203,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 13533
    },
    {
      "epoch": 3.2221031958578825,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2881,
      "step": 13534
    },
    {
      "epoch": 3.222341248586562,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 13535
    },
    {
      "epoch": 3.2225793013152413,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 13536
    },
    {
      "epoch": 3.222817354043921,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 13537
    },
    {
      "epoch": 3.2230554067726,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 13538
    },
    {
      "epoch": 3.2232934595012797,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 13539
    },
    {
      "epoch": 3.223531512229959,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 13540
    },
    {
      "epoch": 3.2237695649586384,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 13541
    },
    {
      "epoch": 3.2240076176873176,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 13542
    },
    {
      "epoch": 3.224245670415997,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 13543
    },
    {
      "epoch": 3.2244837231446764,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 13544
    },
    {
      "epoch": 3.224721775873356,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 13545
    },
    {
      "epoch": 3.224959828602035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 13546
    },
    {
      "epoch": 3.2251978813307147,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 13547
    },
    {
      "epoch": 3.2254359340593943,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 13548
    },
    {
      "epoch": 3.2256739867880735,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 13549
    },
    {
      "epoch": 3.225912039516753,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 13550
    },
    {
      "epoch": 3.2261500922454323,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 13551
    },
    {
      "epoch": 3.226388144974112,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 13552
    },
    {
      "epoch": 3.226626197702791,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 13553
    },
    {
      "epoch": 3.2268642504314706,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 13554
    },
    {
      "epoch": 3.22710230316015,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 13555
    },
    {
      "epoch": 3.2273403558888294,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 13556
    },
    {
      "epoch": 3.227578408617509,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 13557
    },
    {
      "epoch": 3.227816461346188,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 13558
    },
    {
      "epoch": 3.2280545140748678,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 13559
    },
    {
      "epoch": 3.228292566803547,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 13560
    },
    {
      "epoch": 3.2285306195322265,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3867,
      "step": 13561
    },
    {
      "epoch": 3.2287686722609057,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 13562
    },
    {
      "epoch": 3.2290067249895853,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 13563
    },
    {
      "epoch": 3.2292447777182645,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 13564
    },
    {
      "epoch": 3.229482830446944,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 13565
    },
    {
      "epoch": 3.2297208831756232,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 13566
    },
    {
      "epoch": 3.229958935904303,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 13567
    },
    {
      "epoch": 3.230196988632982,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 13568
    },
    {
      "epoch": 3.2304350413616616,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 13569
    },
    {
      "epoch": 3.230673094090341,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 13570
    },
    {
      "epoch": 3.2309111468190204,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 13571
    },
    {
      "epoch": 3.2311491995477,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3899,
      "step": 13572
    },
    {
      "epoch": 3.231387252276379,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 13573
    },
    {
      "epoch": 3.2316253050050587,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 13574
    },
    {
      "epoch": 3.231863357733738,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 13575
    },
    {
      "epoch": 3.2321014104624175,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 13576
    },
    {
      "epoch": 3.2323394631910967,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 13577
    },
    {
      "epoch": 3.2325775159197763,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 13578
    },
    {
      "epoch": 3.2328155686484554,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 13579
    },
    {
      "epoch": 3.233053621377135,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 13580
    },
    {
      "epoch": 3.2332916741058146,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 13581
    },
    {
      "epoch": 3.233529726834494,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4304,
      "step": 13582
    },
    {
      "epoch": 3.2337677795631734,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 13583
    },
    {
      "epoch": 3.2340058322918526,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 13584
    },
    {
      "epoch": 3.234243885020532,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 13585
    },
    {
      "epoch": 3.2344819377492113,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 13586
    },
    {
      "epoch": 3.234719990477891,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 13587
    },
    {
      "epoch": 3.23495804320657,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 13588
    },
    {
      "epoch": 3.2351960959352497,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4769,
      "step": 13589
    },
    {
      "epoch": 3.235434148663929,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 13590
    },
    {
      "epoch": 3.2356722013926085,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 13591
    },
    {
      "epoch": 3.235910254121288,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3596,
      "step": 13592
    },
    {
      "epoch": 3.2361483068499672,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 13593
    },
    {
      "epoch": 3.236386359578647,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 13594
    },
    {
      "epoch": 3.236624412307326,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4251,
      "step": 13595
    },
    {
      "epoch": 3.2368624650360056,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 13596
    },
    {
      "epoch": 3.2371005177646848,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 13597
    },
    {
      "epoch": 3.2373385704933644,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 13598
    },
    {
      "epoch": 3.2375766232220435,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 13599
    },
    {
      "epoch": 3.237814675950723,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3887,
      "step": 13600
    },
    {
      "epoch": 3.237814675950723,
      "eval_runtime": 0.5362,
      "eval_samples_per_second": 1.865,
      "eval_steps_per_second": 1.865,
      "step": 13600
    },
    {
      "epoch": 3.2380527286794023,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 13601
    },
    {
      "epoch": 3.238290781408082,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 13602
    },
    {
      "epoch": 3.238528834136761,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 13603
    },
    {
      "epoch": 3.2387668868654407,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 13604
    },
    {
      "epoch": 3.2390049395941203,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 13605
    },
    {
      "epoch": 3.2392429923227994,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 13606
    },
    {
      "epoch": 3.239481045051479,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 13607
    },
    {
      "epoch": 3.239719097780158,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 13608
    },
    {
      "epoch": 3.239957150508838,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 13609
    },
    {
      "epoch": 3.240195203237517,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 13610
    },
    {
      "epoch": 3.2404332559661966,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 13611
    },
    {
      "epoch": 3.2406713086948757,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2454,
      "step": 13612
    },
    {
      "epoch": 3.2409093614235553,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 13613
    },
    {
      "epoch": 3.241147414152235,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2511,
      "step": 13614
    },
    {
      "epoch": 3.241385466880914,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 13615
    },
    {
      "epoch": 3.2416235196095937,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4709,
      "step": 13616
    },
    {
      "epoch": 3.241861572338273,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3655,
      "step": 13617
    },
    {
      "epoch": 3.2420996250669525,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 13618
    },
    {
      "epoch": 3.2423376777956316,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.3804,
      "step": 13619
    },
    {
      "epoch": 3.2425757305243113,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 13620
    },
    {
      "epoch": 3.2428137832529904,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 13621
    },
    {
      "epoch": 3.24305183598167,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 13622
    },
    {
      "epoch": 3.243289888710349,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 13623
    },
    {
      "epoch": 3.243527941439029,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 13624
    },
    {
      "epoch": 3.243765994167708,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3006,
      "step": 13625
    },
    {
      "epoch": 3.2440040468963875,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 13626
    },
    {
      "epoch": 3.244242099625067,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1506,
      "step": 13627
    },
    {
      "epoch": 3.2444801523537463,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 13628
    },
    {
      "epoch": 3.244718205082426,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 13629
    },
    {
      "epoch": 3.244956257811105,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 13630
    },
    {
      "epoch": 3.2451943105397847,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 13631
    },
    {
      "epoch": 3.245432363268464,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 13632
    },
    {
      "epoch": 3.2456704159971435,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 13633
    },
    {
      "epoch": 3.2459084687258226,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3742,
      "step": 13634
    },
    {
      "epoch": 3.246146521454502,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 13635
    },
    {
      "epoch": 3.2463845741831814,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 13636
    },
    {
      "epoch": 3.246622626911861,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 13637
    },
    {
      "epoch": 3.24686067964054,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.5355,
      "step": 13638
    },
    {
      "epoch": 3.2470987323692198,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 13639
    },
    {
      "epoch": 3.2473367850978994,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 13640
    },
    {
      "epoch": 3.2475748378265785,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 13641
    },
    {
      "epoch": 3.247812890555258,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 13642
    },
    {
      "epoch": 3.2480509432839373,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 13643
    },
    {
      "epoch": 3.248288996012617,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 13644
    },
    {
      "epoch": 3.248527048741296,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 13645
    },
    {
      "epoch": 3.2487651014699757,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 13646
    },
    {
      "epoch": 3.249003154198655,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 13647
    },
    {
      "epoch": 3.2492412069273344,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1909,
      "step": 13648
    },
    {
      "epoch": 3.249479259656014,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 13649
    },
    {
      "epoch": 3.249717312384693,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 13650
    },
    {
      "epoch": 3.249955365113373,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 13651
    },
    {
      "epoch": 3.250193417842052,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 13652
    },
    {
      "epoch": 3.2504314705707316,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 13653
    },
    {
      "epoch": 3.2506695232994107,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 13654
    },
    {
      "epoch": 3.2509075760280903,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4568,
      "step": 13655
    },
    {
      "epoch": 3.2511456287567695,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 13656
    },
    {
      "epoch": 3.251383681485449,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.916,
      "step": 13657
    },
    {
      "epoch": 3.2516217342141283,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 13658
    },
    {
      "epoch": 3.251859786942808,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 13659
    },
    {
      "epoch": 3.252097839671487,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 13660
    },
    {
      "epoch": 3.2523358924001666,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 13661
    },
    {
      "epoch": 3.2525739451288462,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 13662
    },
    {
      "epoch": 3.2528119978575254,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 13663
    },
    {
      "epoch": 3.253050050586205,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 13664
    },
    {
      "epoch": 3.253288103314884,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 13665
    },
    {
      "epoch": 3.2535261560435638,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.769,
      "step": 13666
    },
    {
      "epoch": 3.253764208772243,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 13667
    },
    {
      "epoch": 3.2540022615009225,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 13668
    },
    {
      "epoch": 3.2542403142296017,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 13669
    },
    {
      "epoch": 3.2544783669582813,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 13670
    },
    {
      "epoch": 3.254716419686961,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 13671
    },
    {
      "epoch": 3.25495447241564,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 13672
    },
    {
      "epoch": 3.2551925251443192,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 13673
    },
    {
      "epoch": 3.255430577872999,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 13674
    },
    {
      "epoch": 3.2556686306016784,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 13675
    },
    {
      "epoch": 3.2559066833303576,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 13676
    },
    {
      "epoch": 3.256144736059037,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 13677
    },
    {
      "epoch": 3.2563827887877164,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 13678
    },
    {
      "epoch": 3.256620841516396,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 13679
    },
    {
      "epoch": 3.256858894245075,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 13680
    },
    {
      "epoch": 3.2570969469737547,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 13681
    },
    {
      "epoch": 3.257334999702434,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 13682
    },
    {
      "epoch": 3.2575730524311135,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 13683
    },
    {
      "epoch": 3.257811105159793,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 13684
    },
    {
      "epoch": 3.2580491578884723,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 13685
    },
    {
      "epoch": 3.258287210617152,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 13686
    },
    {
      "epoch": 3.258525263345831,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 13687
    },
    {
      "epoch": 3.2587633160745106,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 13688
    },
    {
      "epoch": 3.25900136880319,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5615,
      "step": 13689
    },
    {
      "epoch": 3.2592394215318694,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 13690
    },
    {
      "epoch": 3.2594774742605486,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 13691
    },
    {
      "epoch": 3.259715526989228,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 13692
    },
    {
      "epoch": 3.2599535797179073,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2365,
      "step": 13693
    },
    {
      "epoch": 3.260191632446587,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3204,
      "step": 13694
    },
    {
      "epoch": 3.260429685175266,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 13695
    },
    {
      "epoch": 3.2606677379039457,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 13696
    },
    {
      "epoch": 3.2609057906326253,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 13697
    },
    {
      "epoch": 3.2611438433613045,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2819,
      "step": 13698
    },
    {
      "epoch": 3.261381896089984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 13699
    },
    {
      "epoch": 3.2616199488186632,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 13700
    },
    {
      "epoch": 3.2616199488186632,
      "eval_runtime": 0.5036,
      "eval_samples_per_second": 1.986,
      "eval_steps_per_second": 1.986,
      "step": 13700
    },
    {
      "epoch": 3.261858001547343,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 13701
    },
    {
      "epoch": 3.262096054276022,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 13702
    },
    {
      "epoch": 3.2623341070047016,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 13703
    },
    {
      "epoch": 3.2625721597333808,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 13704
    },
    {
      "epoch": 3.2628102124620604,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 13705
    },
    {
      "epoch": 3.26304826519074,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 13706
    },
    {
      "epoch": 3.263286317919419,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 13707
    },
    {
      "epoch": 3.2635243706480983,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 13708
    },
    {
      "epoch": 3.263762423376778,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 13709
    },
    {
      "epoch": 3.2640004761054575,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 13710
    },
    {
      "epoch": 3.2642385288341367,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 13711
    },
    {
      "epoch": 3.2644765815628163,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 13712
    },
    {
      "epoch": 3.2647146342914954,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 13713
    },
    {
      "epoch": 3.264952687020175,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 13714
    },
    {
      "epoch": 3.265190739748854,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 13715
    },
    {
      "epoch": 3.265428792477534,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3603,
      "step": 13716
    },
    {
      "epoch": 3.265666845206213,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 13717
    },
    {
      "epoch": 3.2659048979348926,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 13718
    },
    {
      "epoch": 3.266142950663572,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 13719
    },
    {
      "epoch": 3.2663810033922513,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 13720
    },
    {
      "epoch": 3.266619056120931,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 13721
    },
    {
      "epoch": 3.26685710884961,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4129,
      "step": 13722
    },
    {
      "epoch": 3.2670951615782897,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 13723
    },
    {
      "epoch": 3.267333214306969,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7159,
      "step": 13724
    },
    {
      "epoch": 3.2675712670356485,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3136,
      "step": 13725
    },
    {
      "epoch": 3.2678093197643276,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3212,
      "step": 13726
    },
    {
      "epoch": 3.2680473724930073,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.6101,
      "step": 13727
    },
    {
      "epoch": 3.2682854252216864,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 13728
    },
    {
      "epoch": 3.268523477950366,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 13729
    },
    {
      "epoch": 3.268761530679045,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 13730
    },
    {
      "epoch": 3.268999583407725,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 13731
    },
    {
      "epoch": 3.2692376361364044,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 13732
    },
    {
      "epoch": 3.2694756888650836,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 13733
    },
    {
      "epoch": 3.269713741593763,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 13734
    },
    {
      "epoch": 3.2699517943224423,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 13735
    },
    {
      "epoch": 3.270189847051122,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 13736
    },
    {
      "epoch": 3.270427899779801,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 13737
    },
    {
      "epoch": 3.2706659525084807,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 13738
    },
    {
      "epoch": 3.27090400523716,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 13739
    },
    {
      "epoch": 3.2711420579658395,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 13740
    },
    {
      "epoch": 3.271380110694519,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 13741
    },
    {
      "epoch": 3.2716181634231982,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 13742
    },
    {
      "epoch": 3.271856216151878,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 13743
    },
    {
      "epoch": 3.272094268880557,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 13744
    },
    {
      "epoch": 3.2723323216092366,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 13745
    },
    {
      "epoch": 3.2725703743379158,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 13746
    },
    {
      "epoch": 3.2728084270665954,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 13747
    },
    {
      "epoch": 3.2730464797952745,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 13748
    },
    {
      "epoch": 3.273284532523954,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 13749
    },
    {
      "epoch": 3.2735225852526333,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 13750
    },
    {
      "epoch": 3.273760637981313,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 13751
    },
    {
      "epoch": 3.273998690709992,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3238,
      "step": 13752
    },
    {
      "epoch": 3.2742367434386717,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2842,
      "step": 13753
    },
    {
      "epoch": 3.2744747961673513,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 13754
    },
    {
      "epoch": 3.2747128488960304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 13755
    },
    {
      "epoch": 3.27495090162471,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 13756
    },
    {
      "epoch": 3.275188954353389,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3238,
      "step": 13757
    },
    {
      "epoch": 3.275427007082069,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 13758
    },
    {
      "epoch": 3.275665059810748,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3847,
      "step": 13759
    },
    {
      "epoch": 3.2759031125394276,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 13760
    },
    {
      "epoch": 3.2761411652681067,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 13761
    },
    {
      "epoch": 3.2763792179967863,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 13762
    },
    {
      "epoch": 3.276617270725466,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3107,
      "step": 13763
    },
    {
      "epoch": 3.276855323454145,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 13764
    },
    {
      "epoch": 3.2770933761828243,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 13765
    },
    {
      "epoch": 3.277331428911504,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 13766
    },
    {
      "epoch": 3.2775694816401835,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4246,
      "step": 13767
    },
    {
      "epoch": 3.2778075343688626,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4707,
      "step": 13768
    },
    {
      "epoch": 3.2780455870975422,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 13769
    },
    {
      "epoch": 3.2782836398262214,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 13770
    },
    {
      "epoch": 3.278521692554901,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 13771
    },
    {
      "epoch": 3.27875974528358,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 13772
    },
    {
      "epoch": 3.2789977980122598,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 13773
    },
    {
      "epoch": 3.279235850740939,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 13774
    },
    {
      "epoch": 3.2794739034696185,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 13775
    },
    {
      "epoch": 3.279711956198298,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 13776
    },
    {
      "epoch": 3.2799500089269773,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 13777
    },
    {
      "epoch": 3.280188061655657,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 13778
    },
    {
      "epoch": 3.280426114384336,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 13779
    },
    {
      "epoch": 3.2806641671130157,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 13780
    },
    {
      "epoch": 3.280902219841695,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 13781
    },
    {
      "epoch": 3.2811402725703744,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 13782
    },
    {
      "epoch": 3.2813783252990536,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5031,
      "step": 13783
    },
    {
      "epoch": 3.281616378027733,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 13784
    },
    {
      "epoch": 3.2818544307564124,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 13785
    },
    {
      "epoch": 3.282092483485092,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 13786
    },
    {
      "epoch": 3.282330536213771,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 13787
    },
    {
      "epoch": 3.2825685889424507,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 13788
    },
    {
      "epoch": 3.2828066416711303,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4465,
      "step": 13789
    },
    {
      "epoch": 3.2830446943998095,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1415,
      "step": 13790
    },
    {
      "epoch": 3.283282747128489,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4055,
      "step": 13791
    },
    {
      "epoch": 3.2835207998571683,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 13792
    },
    {
      "epoch": 3.283758852585848,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 13793
    },
    {
      "epoch": 3.283996905314527,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 13794
    },
    {
      "epoch": 3.2842349580432066,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 13795
    },
    {
      "epoch": 3.284473010771886,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 13796
    },
    {
      "epoch": 3.2847110635005654,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 13797
    },
    {
      "epoch": 3.284949116229245,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 13798
    },
    {
      "epoch": 3.285187168957924,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 13799
    },
    {
      "epoch": 3.2854252216866033,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 13800
    },
    {
      "epoch": 3.2854252216866033,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 13800
    },
    {
      "epoch": 3.285663274415283,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 13801
    },
    {
      "epoch": 3.2859013271439625,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.467,
      "step": 13802
    },
    {
      "epoch": 3.2861393798726417,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 13803
    },
    {
      "epoch": 3.2863774326013213,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 13804
    },
    {
      "epoch": 3.2866154853300005,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 13805
    },
    {
      "epoch": 3.28685353805868,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 13806
    },
    {
      "epoch": 3.2870915907873592,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 13807
    },
    {
      "epoch": 3.287329643516039,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 13808
    },
    {
      "epoch": 3.287567696244718,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 13809
    },
    {
      "epoch": 3.2878057489733976,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3721,
      "step": 13810
    },
    {
      "epoch": 3.288043801702077,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 13811
    },
    {
      "epoch": 3.2882818544307564,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 13812
    },
    {
      "epoch": 3.288519907159436,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 13813
    },
    {
      "epoch": 3.288757959888115,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 13814
    },
    {
      "epoch": 3.2889960126167948,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 13815
    },
    {
      "epoch": 3.289234065345474,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 13816
    },
    {
      "epoch": 3.2894721180741535,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3372,
      "step": 13817
    },
    {
      "epoch": 3.2897101708028327,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 13818
    },
    {
      "epoch": 3.2899482235315123,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 13819
    },
    {
      "epoch": 3.2901862762601914,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 13820
    },
    {
      "epoch": 3.290424328988871,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 13821
    },
    {
      "epoch": 3.29066238171755,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 13822
    },
    {
      "epoch": 3.29090043444623,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 13823
    },
    {
      "epoch": 3.2911384871749094,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 13824
    },
    {
      "epoch": 3.2913765399035886,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 13825
    },
    {
      "epoch": 3.291614592632268,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 13826
    },
    {
      "epoch": 3.2918526453609473,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3682,
      "step": 13827
    },
    {
      "epoch": 3.292090698089627,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 13828
    },
    {
      "epoch": 3.292328750818306,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 13829
    },
    {
      "epoch": 3.2925668035469857,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 13830
    },
    {
      "epoch": 3.292804856275665,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 13831
    },
    {
      "epoch": 3.2930429090043445,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 13832
    },
    {
      "epoch": 3.293280961733024,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1368,
      "step": 13833
    },
    {
      "epoch": 3.2935190144617033,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2308,
      "step": 13834
    },
    {
      "epoch": 3.293757067190383,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 13835
    },
    {
      "epoch": 3.293995119919062,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 13836
    },
    {
      "epoch": 3.2942331726477416,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 13837
    },
    {
      "epoch": 3.294471225376421,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 13838
    },
    {
      "epoch": 3.2947092781051004,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 13839
    },
    {
      "epoch": 3.2949473308337796,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 13840
    },
    {
      "epoch": 3.295185383562459,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1537,
      "step": 13841
    },
    {
      "epoch": 3.2954234362911383,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 13842
    },
    {
      "epoch": 3.295661489019818,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 13843
    },
    {
      "epoch": 3.295899541748497,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1886,
      "step": 13844
    },
    {
      "epoch": 3.2961375944771767,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 13845
    },
    {
      "epoch": 3.2963756472058563,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 13846
    },
    {
      "epoch": 3.2966136999345355,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 13847
    },
    {
      "epoch": 3.296851752663215,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 13848
    },
    {
      "epoch": 3.2970898053918942,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 13849
    },
    {
      "epoch": 3.297327858120574,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 13850
    },
    {
      "epoch": 3.297565910849253,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 13851
    },
    {
      "epoch": 3.2978039635779326,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 13852
    },
    {
      "epoch": 3.2980420163066118,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 13853
    },
    {
      "epoch": 3.2982800690352914,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 13854
    },
    {
      "epoch": 3.298518121763971,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 13855
    },
    {
      "epoch": 3.29875617449265,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 13856
    },
    {
      "epoch": 3.2989942272213293,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 13857
    },
    {
      "epoch": 3.299232279950009,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 13858
    },
    {
      "epoch": 3.2994703326786885,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 13859
    },
    {
      "epoch": 3.2997083854073677,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6924,
      "step": 13860
    },
    {
      "epoch": 3.2999464381360473,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 13861
    },
    {
      "epoch": 3.3001844908647264,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 13862
    },
    {
      "epoch": 3.300422543593406,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 13863
    },
    {
      "epoch": 3.300660596322085,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 13864
    },
    {
      "epoch": 3.300898649050765,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 13865
    },
    {
      "epoch": 3.301136701779444,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 13866
    },
    {
      "epoch": 3.3013747545081236,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 13867
    },
    {
      "epoch": 3.301612807236803,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 13868
    },
    {
      "epoch": 3.3018508599654823,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 13869
    },
    {
      "epoch": 3.302088912694162,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 13870
    },
    {
      "epoch": 3.302326965422841,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 13871
    },
    {
      "epoch": 3.3025650181515207,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 13872
    },
    {
      "epoch": 3.3028030708802,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 13873
    },
    {
      "epoch": 3.3030411236088795,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 13874
    },
    {
      "epoch": 3.3032791763375586,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 13875
    },
    {
      "epoch": 3.3035172290662382,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 13876
    },
    {
      "epoch": 3.3037552817949174,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 13877
    },
    {
      "epoch": 3.303993334523597,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 13878
    },
    {
      "epoch": 3.304231387252276,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 13879
    },
    {
      "epoch": 3.3044694399809558,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 13880
    },
    {
      "epoch": 3.3047074927096354,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 13881
    },
    {
      "epoch": 3.3049455454383145,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 13882
    },
    {
      "epoch": 3.305183598166994,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 13883
    },
    {
      "epoch": 3.3054216508956733,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 13884
    },
    {
      "epoch": 3.305659703624353,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 13885
    },
    {
      "epoch": 3.305897756353032,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 13886
    },
    {
      "epoch": 3.3061358090817117,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 13887
    },
    {
      "epoch": 3.306373861810391,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 13888
    },
    {
      "epoch": 3.3066119145390704,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 13889
    },
    {
      "epoch": 3.30684996726775,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4298,
      "step": 13890
    },
    {
      "epoch": 3.307088019996429,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 13891
    },
    {
      "epoch": 3.3073260727251084,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 13892
    },
    {
      "epoch": 3.307564125453788,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 13893
    },
    {
      "epoch": 3.3078021781824676,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 13894
    },
    {
      "epoch": 3.3080402309111467,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 13895
    },
    {
      "epoch": 3.3082782836398263,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 13896
    },
    {
      "epoch": 3.3085163363685055,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 13897
    },
    {
      "epoch": 3.308754389097185,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 13898
    },
    {
      "epoch": 3.3089924418258643,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 13899
    },
    {
      "epoch": 3.309230494554544,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3823,
      "step": 13900
    },
    {
      "epoch": 3.309230494554544,
      "eval_runtime": 0.5258,
      "eval_samples_per_second": 1.902,
      "eval_steps_per_second": 1.902,
      "step": 13900
    },
    {
      "epoch": 3.309468547283223,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3904,
      "step": 13901
    },
    {
      "epoch": 3.3097066000119026,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3169,
      "step": 13902
    },
    {
      "epoch": 3.3099446527405822,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 13903
    },
    {
      "epoch": 3.3101827054692614,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 13904
    },
    {
      "epoch": 3.310420758197941,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 13905
    },
    {
      "epoch": 3.31065881092662,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 13906
    },
    {
      "epoch": 3.3108968636553,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 13907
    },
    {
      "epoch": 3.311134916383979,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 13908
    },
    {
      "epoch": 3.3113729691126585,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 13909
    },
    {
      "epoch": 3.3116110218413377,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 13910
    },
    {
      "epoch": 3.3118490745700173,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 13911
    },
    {
      "epoch": 3.3120871272986965,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4071,
      "step": 13912
    },
    {
      "epoch": 3.312325180027376,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 13913
    },
    {
      "epoch": 3.3125632327560552,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 13914
    },
    {
      "epoch": 3.312801285484735,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 13915
    },
    {
      "epoch": 3.3130393382134145,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2526,
      "step": 13916
    },
    {
      "epoch": 3.3132773909420936,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 13917
    },
    {
      "epoch": 3.313515443670773,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 13918
    },
    {
      "epoch": 3.3137534963994524,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 13919
    },
    {
      "epoch": 3.313991549128132,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 13920
    },
    {
      "epoch": 3.314229601856811,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 13921
    },
    {
      "epoch": 3.3144676545854908,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 13922
    },
    {
      "epoch": 3.31470570731417,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 13923
    },
    {
      "epoch": 3.3149437600428495,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.304,
      "step": 13924
    },
    {
      "epoch": 3.315181812771529,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 13925
    },
    {
      "epoch": 3.3154198655002083,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 13926
    },
    {
      "epoch": 3.315657918228888,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 13927
    },
    {
      "epoch": 3.315895970957567,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4181,
      "step": 13928
    },
    {
      "epoch": 3.3161340236862467,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 13929
    },
    {
      "epoch": 3.316372076414926,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 13930
    },
    {
      "epoch": 3.3166101291436054,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 13931
    },
    {
      "epoch": 3.3168481818722846,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 13932
    },
    {
      "epoch": 3.317086234600964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 13933
    },
    {
      "epoch": 3.3173242873296434,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 13934
    },
    {
      "epoch": 3.317562340058323,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 13935
    },
    {
      "epoch": 3.317800392787002,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 13936
    },
    {
      "epoch": 3.3180384455156817,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 13937
    },
    {
      "epoch": 3.3182764982443613,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 13938
    },
    {
      "epoch": 3.3185145509730405,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 13939
    },
    {
      "epoch": 3.31875260370172,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 13940
    },
    {
      "epoch": 3.3189906564303993,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 13941
    },
    {
      "epoch": 3.319228709159079,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 13942
    },
    {
      "epoch": 3.319466761887758,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 13943
    },
    {
      "epoch": 3.3197048146164376,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 13944
    },
    {
      "epoch": 3.319942867345117,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.238,
      "step": 13945
    },
    {
      "epoch": 3.3201809200737964,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 13946
    },
    {
      "epoch": 3.3204189728024756,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 13947
    },
    {
      "epoch": 3.320657025531155,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 13948
    },
    {
      "epoch": 3.3208950782598343,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 13949
    },
    {
      "epoch": 3.321133130988514,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 13950
    },
    {
      "epoch": 3.3213711837171935,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 13951
    },
    {
      "epoch": 3.3216092364458727,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 13952
    },
    {
      "epoch": 3.3218472891745523,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 13953
    },
    {
      "epoch": 3.3220853419032315,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 13954
    },
    {
      "epoch": 3.322323394631911,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 13955
    },
    {
      "epoch": 3.3225614473605902,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 13956
    },
    {
      "epoch": 3.32279950008927,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 13957
    },
    {
      "epoch": 3.323037552817949,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 13958
    },
    {
      "epoch": 3.3232756055466286,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 13959
    },
    {
      "epoch": 3.323513658275308,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 13960
    },
    {
      "epoch": 3.3237517110039874,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 13961
    },
    {
      "epoch": 3.323989763732667,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 13962
    },
    {
      "epoch": 3.324227816461346,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 13963
    },
    {
      "epoch": 3.3244658691900257,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4087,
      "step": 13964
    },
    {
      "epoch": 3.324703921918705,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4192,
      "step": 13965
    },
    {
      "epoch": 3.3249419746473845,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 13966
    },
    {
      "epoch": 3.3251800273760637,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 13967
    },
    {
      "epoch": 3.3254180801047433,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 13968
    },
    {
      "epoch": 3.3256561328334224,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 13969
    },
    {
      "epoch": 3.325894185562102,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 13970
    },
    {
      "epoch": 3.326132238290781,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 13971
    },
    {
      "epoch": 3.326370291019461,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 13972
    },
    {
      "epoch": 3.3266083437481404,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 13973
    },
    {
      "epoch": 3.3268463964768196,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 13974
    },
    {
      "epoch": 3.327084449205499,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 13975
    },
    {
      "epoch": 3.3273225019341783,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 13976
    },
    {
      "epoch": 3.327560554662858,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 13977
    },
    {
      "epoch": 3.327798607391537,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3976,
      "step": 13978
    },
    {
      "epoch": 3.3280366601202167,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 13979
    },
    {
      "epoch": 3.328274712848896,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 13980
    },
    {
      "epoch": 3.3285127655775755,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 13981
    },
    {
      "epoch": 3.328750818306255,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 13982
    },
    {
      "epoch": 3.3289888710349342,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 13983
    },
    {
      "epoch": 3.3292269237636134,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 13984
    },
    {
      "epoch": 3.329464976492293,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 13985
    },
    {
      "epoch": 3.3297030292209726,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7164,
      "step": 13986
    },
    {
      "epoch": 3.3299410819496518,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 13987
    },
    {
      "epoch": 3.3301791346783314,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4101,
      "step": 13988
    },
    {
      "epoch": 3.3304171874070105,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 13989
    },
    {
      "epoch": 3.33065524013569,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 13990
    },
    {
      "epoch": 3.3308932928643693,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 13991
    },
    {
      "epoch": 3.331131345593049,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 13992
    },
    {
      "epoch": 3.331369398321728,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 13993
    },
    {
      "epoch": 3.3316074510504077,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.6046,
      "step": 13994
    },
    {
      "epoch": 3.3318455037790873,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 13995
    },
    {
      "epoch": 3.3320835565077664,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 13996
    },
    {
      "epoch": 3.332321609236446,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3097,
      "step": 13997
    },
    {
      "epoch": 3.332559661965125,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 13998
    },
    {
      "epoch": 3.332797714693805,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 13999
    },
    {
      "epoch": 3.333035767422484,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 14000
    },
    {
      "epoch": 3.333035767422484,
      "eval_runtime": 0.5127,
      "eval_samples_per_second": 1.95,
      "eval_steps_per_second": 1.95,
      "step": 14000
    },
    {
      "epoch": 3.3332738201511636,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 14001
    },
    {
      "epoch": 3.3335118728798427,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 14002
    },
    {
      "epoch": 3.3337499256085223,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 14003
    },
    {
      "epoch": 3.3339879783372015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 14004
    },
    {
      "epoch": 3.334226031065881,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.25,
      "step": 14005
    },
    {
      "epoch": 3.3344640837945603,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 14006
    },
    {
      "epoch": 3.33470213652324,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 14007
    },
    {
      "epoch": 3.3349401892519195,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3521,
      "step": 14008
    },
    {
      "epoch": 3.3351782419805986,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 14009
    },
    {
      "epoch": 3.3354162947092783,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 14010
    },
    {
      "epoch": 3.3356543474379574,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 14011
    },
    {
      "epoch": 3.335892400166637,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 14012
    },
    {
      "epoch": 3.336130452895316,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 14013
    },
    {
      "epoch": 3.336368505623996,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1647,
      "step": 14014
    },
    {
      "epoch": 3.336606558352675,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 14015
    },
    {
      "epoch": 3.3368446110813546,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5541,
      "step": 14016
    },
    {
      "epoch": 3.337082663810034,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 14017
    },
    {
      "epoch": 3.3373207165387133,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 14018
    },
    {
      "epoch": 3.3375587692673925,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 14019
    },
    {
      "epoch": 3.337796821996072,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 14020
    },
    {
      "epoch": 3.3380348747247517,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4981,
      "step": 14021
    },
    {
      "epoch": 3.338272927453431,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 14022
    },
    {
      "epoch": 3.3385109801821105,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 14023
    },
    {
      "epoch": 3.3387490329107896,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 14024
    },
    {
      "epoch": 3.338987085639469,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3836,
      "step": 14025
    },
    {
      "epoch": 3.3392251383681484,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 14026
    },
    {
      "epoch": 3.339463191096828,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 14027
    },
    {
      "epoch": 3.339701243825507,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 14028
    },
    {
      "epoch": 3.3399392965541868,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 14029
    },
    {
      "epoch": 3.3401773492828664,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 14030
    },
    {
      "epoch": 3.3404154020115455,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 14031
    },
    {
      "epoch": 3.340653454740225,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 14032
    },
    {
      "epoch": 3.3408915074689043,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 14033
    },
    {
      "epoch": 3.341129560197584,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 14034
    },
    {
      "epoch": 3.341367612926263,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 14035
    },
    {
      "epoch": 3.3416056656549427,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 14036
    },
    {
      "epoch": 3.341843718383622,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 14037
    },
    {
      "epoch": 3.3420817711123014,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 14038
    },
    {
      "epoch": 3.3423198238409806,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 14039
    },
    {
      "epoch": 3.34255787656966,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 14040
    },
    {
      "epoch": 3.3427959292983394,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 14041
    },
    {
      "epoch": 3.343033982027019,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 14042
    },
    {
      "epoch": 3.3432720347556986,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 14043
    },
    {
      "epoch": 3.3435100874843777,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 14044
    },
    {
      "epoch": 3.3437481402130573,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 14045
    },
    {
      "epoch": 3.3439861929417365,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.438,
      "step": 14046
    },
    {
      "epoch": 3.344224245670416,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3117,
      "step": 14047
    },
    {
      "epoch": 3.3444622983990953,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3385,
      "step": 14048
    },
    {
      "epoch": 3.344700351127775,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 14049
    },
    {
      "epoch": 3.344938403856454,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 14050
    },
    {
      "epoch": 3.3451764565851336,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 14051
    },
    {
      "epoch": 3.3454145093138132,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.5118,
      "step": 14052
    },
    {
      "epoch": 3.3456525620424924,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 14053
    },
    {
      "epoch": 3.345890614771172,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2283,
      "step": 14054
    },
    {
      "epoch": 3.346128667499851,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 14055
    },
    {
      "epoch": 3.3463667202285308,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 14056
    },
    {
      "epoch": 3.34660477295721,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 14057
    },
    {
      "epoch": 3.3468428256858895,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.735,
      "step": 14058
    },
    {
      "epoch": 3.3470808784145687,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 14059
    },
    {
      "epoch": 3.3473189311432483,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 14060
    },
    {
      "epoch": 3.3475569838719275,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 14061
    },
    {
      "epoch": 3.347795036600607,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 14062
    },
    {
      "epoch": 3.3480330893292862,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 14063
    },
    {
      "epoch": 3.348271142057966,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2455,
      "step": 14064
    },
    {
      "epoch": 3.3485091947866454,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 14065
    },
    {
      "epoch": 3.3487472475153246,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 14066
    },
    {
      "epoch": 3.348985300244004,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 14067
    },
    {
      "epoch": 3.3492233529726834,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 14068
    },
    {
      "epoch": 3.349461405701363,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 14069
    },
    {
      "epoch": 3.349699458430042,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 14070
    },
    {
      "epoch": 3.3499375111587217,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 14071
    },
    {
      "epoch": 3.350175563887401,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 14072
    },
    {
      "epoch": 3.3504136166160805,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5446,
      "step": 14073
    },
    {
      "epoch": 3.35065166934476,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 14074
    },
    {
      "epoch": 3.3508897220734393,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 14075
    },
    {
      "epoch": 3.3511277748021184,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3548,
      "step": 14076
    },
    {
      "epoch": 3.351365827530798,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 14077
    },
    {
      "epoch": 3.3516038802594776,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 14078
    },
    {
      "epoch": 3.351841932988157,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 14079
    },
    {
      "epoch": 3.3520799857168364,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 14080
    },
    {
      "epoch": 3.3523180384455156,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 14081
    },
    {
      "epoch": 3.352556091174195,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9098,
      "step": 14082
    },
    {
      "epoch": 3.3527941439028743,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 14083
    },
    {
      "epoch": 3.353032196631554,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 14084
    },
    {
      "epoch": 3.353270249360233,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 14085
    },
    {
      "epoch": 3.3535083020889127,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 14086
    },
    {
      "epoch": 3.3537463548175923,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 14087
    },
    {
      "epoch": 3.3539844075462715,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 14088
    },
    {
      "epoch": 3.354222460274951,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 14089
    },
    {
      "epoch": 3.3544605130036302,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 14090
    },
    {
      "epoch": 3.35469856573231,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 14091
    },
    {
      "epoch": 3.354936618460989,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 14092
    },
    {
      "epoch": 3.3551746711896686,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 14093
    },
    {
      "epoch": 3.3554127239183478,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 14094
    },
    {
      "epoch": 3.3556507766470274,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 14095
    },
    {
      "epoch": 3.3558888293757065,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 14096
    },
    {
      "epoch": 3.356126882104386,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 14097
    },
    {
      "epoch": 3.3563649348330653,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 14098
    },
    {
      "epoch": 3.356602987561745,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.3524,
      "step": 14099
    },
    {
      "epoch": 3.3568410402904245,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 14100
    },
    {
      "epoch": 3.3568410402904245,
      "eval_runtime": 0.5179,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 14100
    },
    {
      "epoch": 3.3570790930191037,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4678,
      "step": 14101
    },
    {
      "epoch": 3.3573171457477833,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 14102
    },
    {
      "epoch": 3.3575551984764624,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 14103
    },
    {
      "epoch": 3.357793251205142,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 14104
    },
    {
      "epoch": 3.358031303933821,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2772,
      "step": 14105
    },
    {
      "epoch": 3.358269356662501,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 14106
    },
    {
      "epoch": 3.35850740939118,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 14107
    },
    {
      "epoch": 3.3587454621198596,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 14108
    },
    {
      "epoch": 3.358983514848539,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 14109
    },
    {
      "epoch": 3.3592215675772183,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 14110
    },
    {
      "epoch": 3.3594596203058975,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 14111
    },
    {
      "epoch": 3.359697673034577,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 14112
    },
    {
      "epoch": 3.3599357257632567,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 14113
    },
    {
      "epoch": 3.360173778491936,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 14114
    },
    {
      "epoch": 3.3604118312206155,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 14115
    },
    {
      "epoch": 3.3606498839492946,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 14116
    },
    {
      "epoch": 3.3608879366779743,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 14117
    },
    {
      "epoch": 3.3611259894066534,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 14118
    },
    {
      "epoch": 3.361364042135333,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 14119
    },
    {
      "epoch": 3.361602094864012,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 14120
    },
    {
      "epoch": 3.361840147592692,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 14121
    },
    {
      "epoch": 3.3620782003213714,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 14122
    },
    {
      "epoch": 3.3623162530500506,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 14123
    },
    {
      "epoch": 3.36255430577873,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 14124
    },
    {
      "epoch": 3.3627923585074093,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3672,
      "step": 14125
    },
    {
      "epoch": 3.363030411236089,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.373,
      "step": 14126
    },
    {
      "epoch": 3.363268463964768,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 14127
    },
    {
      "epoch": 3.3635065166934477,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 14128
    },
    {
      "epoch": 3.363744569422127,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 14129
    },
    {
      "epoch": 3.3639826221508065,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 14130
    },
    {
      "epoch": 3.3642206748794856,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 14131
    },
    {
      "epoch": 3.3644587276081652,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 14132
    },
    {
      "epoch": 3.3646967803368444,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 14133
    },
    {
      "epoch": 3.364934833065524,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 14134
    },
    {
      "epoch": 3.3651728857942036,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 14135
    },
    {
      "epoch": 3.3654109385228828,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 14136
    },
    {
      "epoch": 3.3656489912515624,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 14137
    },
    {
      "epoch": 3.3658870439802415,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 14138
    },
    {
      "epoch": 3.366125096708921,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 14139
    },
    {
      "epoch": 3.3663631494376003,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 14140
    },
    {
      "epoch": 3.36660120216628,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 14141
    },
    {
      "epoch": 3.366839254894959,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 14142
    },
    {
      "epoch": 3.3670773076236387,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 14143
    },
    {
      "epoch": 3.3673153603523183,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 14144
    },
    {
      "epoch": 3.3675534130809974,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2744,
      "step": 14145
    },
    {
      "epoch": 3.367791465809677,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 14146
    },
    {
      "epoch": 3.368029518538356,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 14147
    },
    {
      "epoch": 3.368267571267036,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 14148
    },
    {
      "epoch": 3.368505623995715,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 14149
    },
    {
      "epoch": 3.3687436767243946,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 14150
    },
    {
      "epoch": 3.3689817294530737,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 14151
    },
    {
      "epoch": 3.3692197821817533,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 14152
    },
    {
      "epoch": 3.3694578349104325,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 14153
    },
    {
      "epoch": 3.369695887639112,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 14154
    },
    {
      "epoch": 3.3699339403677913,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2577,
      "step": 14155
    },
    {
      "epoch": 3.370171993096471,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 14156
    },
    {
      "epoch": 3.3704100458251505,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 14157
    },
    {
      "epoch": 3.3706480985538296,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4693,
      "step": 14158
    },
    {
      "epoch": 3.3708861512825092,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 14159
    },
    {
      "epoch": 3.3711242040111884,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 14160
    },
    {
      "epoch": 3.371362256739868,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 14161
    },
    {
      "epoch": 3.371600309468547,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 14162
    },
    {
      "epoch": 3.3718383621972268,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 14163
    },
    {
      "epoch": 3.372076414925906,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 14164
    },
    {
      "epoch": 3.3723144676545855,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 14165
    },
    {
      "epoch": 3.372552520383265,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 14166
    },
    {
      "epoch": 3.3727905731119443,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 14167
    },
    {
      "epoch": 3.3730286258406235,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 14168
    },
    {
      "epoch": 3.373266678569303,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 14169
    },
    {
      "epoch": 3.3735047312979827,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 14170
    },
    {
      "epoch": 3.373742784026662,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3293,
      "step": 14171
    },
    {
      "epoch": 3.3739808367553414,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 14172
    },
    {
      "epoch": 3.3742188894840206,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 14173
    },
    {
      "epoch": 3.3744569422127,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 14174
    },
    {
      "epoch": 3.3746949949413794,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 14175
    },
    {
      "epoch": 3.374933047670059,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 14176
    },
    {
      "epoch": 3.375171100398738,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4149,
      "step": 14177
    },
    {
      "epoch": 3.3754091531274177,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 14178
    },
    {
      "epoch": 3.3756472058560973,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 14179
    },
    {
      "epoch": 3.3758852585847765,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 14180
    },
    {
      "epoch": 3.376123311313456,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 14181
    },
    {
      "epoch": 3.3763613640421353,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 14182
    },
    {
      "epoch": 3.376599416770815,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4598,
      "step": 14183
    },
    {
      "epoch": 3.376837469499494,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 14184
    },
    {
      "epoch": 3.3770755222281736,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 14185
    },
    {
      "epoch": 3.377313574956853,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 14186
    },
    {
      "epoch": 3.3775516276855324,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.6887,
      "step": 14187
    },
    {
      "epoch": 3.3777896804142116,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 14188
    },
    {
      "epoch": 3.378027733142891,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 14189
    },
    {
      "epoch": 3.3782657858715703,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 14190
    },
    {
      "epoch": 3.37850383860025,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 14191
    },
    {
      "epoch": 3.3787418913289295,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 14192
    },
    {
      "epoch": 3.3789799440576087,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 14193
    },
    {
      "epoch": 3.3792179967862883,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 14194
    },
    {
      "epoch": 3.3794560495149675,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 14195
    },
    {
      "epoch": 3.379694102243647,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 14196
    },
    {
      "epoch": 3.3799321549723262,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 14197
    },
    {
      "epoch": 3.380170207701006,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 14198
    },
    {
      "epoch": 3.380408260429685,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 14199
    },
    {
      "epoch": 3.3806463131583646,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3698,
      "step": 14200
    },
    {
      "epoch": 3.3806463131583646,
      "eval_runtime": 0.5302,
      "eval_samples_per_second": 1.886,
      "eval_steps_per_second": 1.886,
      "step": 14200
    },
    {
      "epoch": 3.380884365887044,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 14201
    },
    {
      "epoch": 3.3811224186157234,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 14202
    },
    {
      "epoch": 3.3813604713444025,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 14203
    },
    {
      "epoch": 3.381598524073082,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 14204
    },
    {
      "epoch": 3.3818365768017618,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 14205
    },
    {
      "epoch": 3.382074629530441,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 14206
    },
    {
      "epoch": 3.3823126822591205,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2693,
      "step": 14207
    },
    {
      "epoch": 3.3825507349877997,
      "grad_norm": 7.84375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 14208
    },
    {
      "epoch": 3.3827887877164793,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 14209
    },
    {
      "epoch": 3.3830268404451584,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 14210
    },
    {
      "epoch": 3.383264893173838,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 14211
    },
    {
      "epoch": 3.383502945902517,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3985,
      "step": 14212
    },
    {
      "epoch": 3.383740998631197,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 14213
    },
    {
      "epoch": 3.3839790513598764,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 14214
    },
    {
      "epoch": 3.3842171040885556,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 14215
    },
    {
      "epoch": 3.384455156817235,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 14216
    },
    {
      "epoch": 3.3846932095459143,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 14217
    },
    {
      "epoch": 3.384931262274594,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 14218
    },
    {
      "epoch": 3.385169315003273,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 14219
    },
    {
      "epoch": 3.3854073677319527,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 14220
    },
    {
      "epoch": 3.385645420460632,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 14221
    },
    {
      "epoch": 3.3858834731893115,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 14222
    },
    {
      "epoch": 3.3861215259179906,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 14223
    },
    {
      "epoch": 3.3863595786466703,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3339,
      "step": 14224
    },
    {
      "epoch": 3.3865976313753494,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 14225
    },
    {
      "epoch": 3.386835684104029,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 14226
    },
    {
      "epoch": 3.3870737368327086,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 14227
    },
    {
      "epoch": 3.387311789561388,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 14228
    },
    {
      "epoch": 3.3875498422900674,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 14229
    },
    {
      "epoch": 3.3877878950187466,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 14230
    },
    {
      "epoch": 3.388025947747426,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 14231
    },
    {
      "epoch": 3.3882640004761053,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 14232
    },
    {
      "epoch": 3.388502053204785,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 14233
    },
    {
      "epoch": 3.388740105933464,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 14234
    },
    {
      "epoch": 3.3889781586621437,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 14235
    },
    {
      "epoch": 3.3892162113908233,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 14236
    },
    {
      "epoch": 3.3894542641195025,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 14237
    },
    {
      "epoch": 3.389692316848182,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 14238
    },
    {
      "epoch": 3.3899303695768612,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 14239
    },
    {
      "epoch": 3.390168422305541,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4147,
      "step": 14240
    },
    {
      "epoch": 3.39040647503422,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.414,
      "step": 14241
    },
    {
      "epoch": 3.3906445277628996,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 14242
    },
    {
      "epoch": 3.3908825804915788,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.5781,
      "step": 14243
    },
    {
      "epoch": 3.3911206332202584,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4909,
      "step": 14244
    },
    {
      "epoch": 3.3913586859489375,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 14245
    },
    {
      "epoch": 3.391596738677617,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 14246
    },
    {
      "epoch": 3.3918347914062963,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 14247
    },
    {
      "epoch": 3.392072844134976,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4852,
      "step": 14248
    },
    {
      "epoch": 3.3923108968636555,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 14249
    },
    {
      "epoch": 3.3925489495923347,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 14250
    },
    {
      "epoch": 3.3927870023210143,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 14251
    },
    {
      "epoch": 3.3930250550496934,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 14252
    },
    {
      "epoch": 3.393263107778373,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 14253
    },
    {
      "epoch": 3.393501160507052,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 14254
    },
    {
      "epoch": 3.393739213235732,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 14255
    },
    {
      "epoch": 3.393977265964411,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 14256
    },
    {
      "epoch": 3.3942153186930906,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 14257
    },
    {
      "epoch": 3.39445337142177,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 14258
    },
    {
      "epoch": 3.3946914241504493,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3213,
      "step": 14259
    },
    {
      "epoch": 3.3949294768791285,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 14260
    },
    {
      "epoch": 3.395167529607808,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.284,
      "step": 14261
    },
    {
      "epoch": 3.3954055823364877,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 14262
    },
    {
      "epoch": 3.395643635065167,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 14263
    },
    {
      "epoch": 3.3958816877938465,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 14264
    },
    {
      "epoch": 3.3961197405225256,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 14265
    },
    {
      "epoch": 3.3963577932512052,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 14266
    },
    {
      "epoch": 3.3965958459798844,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 14267
    },
    {
      "epoch": 3.396833898708564,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 14268
    },
    {
      "epoch": 3.397071951437243,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 14269
    },
    {
      "epoch": 3.3973100041659228,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 14270
    },
    {
      "epoch": 3.3975480568946024,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.37,
      "step": 14271
    },
    {
      "epoch": 3.3977861096232815,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 14272
    },
    {
      "epoch": 3.398024162351961,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 14273
    },
    {
      "epoch": 3.3982622150806403,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 14274
    },
    {
      "epoch": 3.39850026780932,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 14275
    },
    {
      "epoch": 3.398738320537999,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 14276
    },
    {
      "epoch": 3.3989763732666787,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 14277
    },
    {
      "epoch": 3.399214425995358,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 14278
    },
    {
      "epoch": 3.3994524787240374,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 14279
    },
    {
      "epoch": 3.3996905314527166,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 14280
    },
    {
      "epoch": 3.399928584181396,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3258,
      "step": 14281
    },
    {
      "epoch": 3.4001666369100754,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 14282
    },
    {
      "epoch": 3.400404689638755,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 14283
    },
    {
      "epoch": 3.4006427423674346,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 14284
    },
    {
      "epoch": 3.4008807950961137,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 14285
    },
    {
      "epoch": 3.4011188478247933,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.5528,
      "step": 14286
    },
    {
      "epoch": 3.4013569005534725,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 14287
    },
    {
      "epoch": 3.401594953282152,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 14288
    },
    {
      "epoch": 3.4018330060108313,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 14289
    },
    {
      "epoch": 3.402071058739511,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 14290
    },
    {
      "epoch": 3.40230911146819,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 14291
    },
    {
      "epoch": 3.4025471641968696,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 14292
    },
    {
      "epoch": 3.4027852169255492,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 14293
    },
    {
      "epoch": 3.4030232696542284,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 14294
    },
    {
      "epoch": 3.4032613223829076,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 14295
    },
    {
      "epoch": 3.403499375111587,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3982,
      "step": 14296
    },
    {
      "epoch": 3.403737427840267,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 14297
    },
    {
      "epoch": 3.403975480568946,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3209,
      "step": 14298
    },
    {
      "epoch": 3.4042135332976255,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 14299
    },
    {
      "epoch": 3.4044515860263047,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 14300
    },
    {
      "epoch": 3.4044515860263047,
      "eval_runtime": 0.5268,
      "eval_samples_per_second": 1.898,
      "eval_steps_per_second": 1.898,
      "step": 14300
    },
    {
      "epoch": 3.4046896387549843,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 14301
    },
    {
      "epoch": 3.4049276914836635,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 14302
    },
    {
      "epoch": 3.405165744212343,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 14303
    },
    {
      "epoch": 3.4054037969410222,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 14304
    },
    {
      "epoch": 3.405641849669702,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 14305
    },
    {
      "epoch": 3.4058799023983815,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 14306
    },
    {
      "epoch": 3.4061179551270606,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 14307
    },
    {
      "epoch": 3.40635600785574,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 14308
    },
    {
      "epoch": 3.4065940605844194,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 14309
    },
    {
      "epoch": 3.406832113313099,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 14310
    },
    {
      "epoch": 3.407070166041778,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 14311
    },
    {
      "epoch": 3.4073082187704578,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 14312
    },
    {
      "epoch": 3.407546271499137,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 14313
    },
    {
      "epoch": 3.4077843242278165,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 14314
    },
    {
      "epoch": 3.4080223769564957,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 14315
    },
    {
      "epoch": 3.4082604296851753,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 14316
    },
    {
      "epoch": 3.4084984824138544,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.6908,
      "step": 14317
    },
    {
      "epoch": 3.408736535142534,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 14318
    },
    {
      "epoch": 3.4089745878712137,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3171,
      "step": 14319
    },
    {
      "epoch": 3.409212640599893,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 14320
    },
    {
      "epoch": 3.4094506933285724,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 14321
    },
    {
      "epoch": 3.4096887460572516,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2901,
      "step": 14322
    },
    {
      "epoch": 3.409926798785931,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 14323
    },
    {
      "epoch": 3.4101648515146104,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 14324
    },
    {
      "epoch": 3.41040290424329,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 14325
    },
    {
      "epoch": 3.410640956971969,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 14326
    },
    {
      "epoch": 3.4108790097006487,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 14327
    },
    {
      "epoch": 3.4111170624293283,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 14328
    },
    {
      "epoch": 3.4113551151580075,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 14329
    },
    {
      "epoch": 3.411593167886687,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 14330
    },
    {
      "epoch": 3.4118312206153663,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 14331
    },
    {
      "epoch": 3.412069273344046,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 14332
    },
    {
      "epoch": 3.412307326072725,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 14333
    },
    {
      "epoch": 3.4125453788014046,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 14334
    },
    {
      "epoch": 3.412783431530084,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 14335
    },
    {
      "epoch": 3.4130214842587634,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 14336
    },
    {
      "epoch": 3.4132595369874426,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7305,
      "step": 14337
    },
    {
      "epoch": 3.413497589716122,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4117,
      "step": 14338
    },
    {
      "epoch": 3.4137356424448013,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 14339
    },
    {
      "epoch": 3.413973695173481,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 14340
    },
    {
      "epoch": 3.4142117479021605,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 14341
    },
    {
      "epoch": 3.4144498006308397,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 14342
    },
    {
      "epoch": 3.4146878533595193,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 14343
    },
    {
      "epoch": 3.4149259060881985,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 14344
    },
    {
      "epoch": 3.415163958816878,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 14345
    },
    {
      "epoch": 3.4154020115455572,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 14346
    },
    {
      "epoch": 3.415640064274237,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 14347
    },
    {
      "epoch": 3.415878117002916,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 14348
    },
    {
      "epoch": 3.4161161697315956,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 14349
    },
    {
      "epoch": 3.4163542224602748,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 14350
    },
    {
      "epoch": 3.4165922751889544,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 14351
    },
    {
      "epoch": 3.4168303279176335,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 14352
    },
    {
      "epoch": 3.417068380646313,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 14353
    },
    {
      "epoch": 3.4173064333749927,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 14354
    },
    {
      "epoch": 3.417544486103672,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 14355
    },
    {
      "epoch": 3.4177825388323515,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 14356
    },
    {
      "epoch": 3.4180205915610307,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 14357
    },
    {
      "epoch": 3.4182586442897103,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 14358
    },
    {
      "epoch": 3.4184966970183894,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 14359
    },
    {
      "epoch": 3.418734749747069,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 14360
    },
    {
      "epoch": 3.418972802475748,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 14361
    },
    {
      "epoch": 3.419210855204428,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 14362
    },
    {
      "epoch": 3.4194489079331074,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 14363
    },
    {
      "epoch": 3.4196869606617866,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 14364
    },
    {
      "epoch": 3.419925013390466,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3364,
      "step": 14365
    },
    {
      "epoch": 3.4201630661191453,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 14366
    },
    {
      "epoch": 3.420401118847825,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 14367
    },
    {
      "epoch": 3.420639171576504,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 14368
    },
    {
      "epoch": 3.4208772243051837,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 14369
    },
    {
      "epoch": 3.421115277033863,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 14370
    },
    {
      "epoch": 3.4213533297625425,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 14371
    },
    {
      "epoch": 3.4215913824912216,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 14372
    },
    {
      "epoch": 3.4218294352199012,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 14373
    },
    {
      "epoch": 3.4220674879485804,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5188,
      "step": 14374
    },
    {
      "epoch": 3.42230554067726,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 14375
    },
    {
      "epoch": 3.4225435934059396,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3236,
      "step": 14376
    },
    {
      "epoch": 3.4227816461346188,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 14377
    },
    {
      "epoch": 3.4230196988632984,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 14378
    },
    {
      "epoch": 3.4232577515919775,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 14379
    },
    {
      "epoch": 3.423495804320657,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.227,
      "step": 14380
    },
    {
      "epoch": 3.4237338570493363,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 14381
    },
    {
      "epoch": 3.423971909778016,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 14382
    },
    {
      "epoch": 3.424209962506695,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 14383
    },
    {
      "epoch": 3.4244480152353747,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5074,
      "step": 14384
    },
    {
      "epoch": 3.4246860679640543,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 14385
    },
    {
      "epoch": 3.4249241206927334,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 14386
    },
    {
      "epoch": 3.4251621734214126,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 14387
    },
    {
      "epoch": 3.425400226150092,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 14388
    },
    {
      "epoch": 3.425638278878772,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 14389
    },
    {
      "epoch": 3.425876331607451,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 14390
    },
    {
      "epoch": 3.4261143843361306,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 14391
    },
    {
      "epoch": 3.4263524370648097,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 14392
    },
    {
      "epoch": 3.4265904897934893,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 14393
    },
    {
      "epoch": 3.4268285425221685,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 14394
    },
    {
      "epoch": 3.427066595250848,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 14395
    },
    {
      "epoch": 3.4273046479795273,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3778,
      "step": 14396
    },
    {
      "epoch": 3.427542700708207,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3898,
      "step": 14397
    },
    {
      "epoch": 3.4277807534368865,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 14398
    },
    {
      "epoch": 3.4280188061655656,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 14399
    },
    {
      "epoch": 3.4282568588942453,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 14400
    },
    {
      "epoch": 3.4282568588942453,
      "eval_runtime": 0.5219,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 14400
    },
    {
      "epoch": 3.4284949116229244,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 14401
    },
    {
      "epoch": 3.428732964351604,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 14402
    },
    {
      "epoch": 3.428971017080283,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 14403
    },
    {
      "epoch": 3.429209069808963,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 14404
    },
    {
      "epoch": 3.429447122537642,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 14405
    },
    {
      "epoch": 3.4296851752663216,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 14406
    },
    {
      "epoch": 3.4299232279950007,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 14407
    },
    {
      "epoch": 3.4301612807236803,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4774,
      "step": 14408
    },
    {
      "epoch": 3.4303993334523595,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 14409
    },
    {
      "epoch": 3.430637386181039,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.5197,
      "step": 14410
    },
    {
      "epoch": 3.4308754389097187,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 14411
    },
    {
      "epoch": 3.431113491638398,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 14412
    },
    {
      "epoch": 3.4313515443670775,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 14413
    },
    {
      "epoch": 3.4315895970957566,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 14414
    },
    {
      "epoch": 3.431827649824436,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 14415
    },
    {
      "epoch": 3.4320657025531154,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4778,
      "step": 14416
    },
    {
      "epoch": 3.432303755281795,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 14417
    },
    {
      "epoch": 3.432541808010474,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 14418
    },
    {
      "epoch": 3.4327798607391538,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 14419
    },
    {
      "epoch": 3.4330179134678334,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2944,
      "step": 14420
    },
    {
      "epoch": 3.4332559661965125,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2943,
      "step": 14421
    },
    {
      "epoch": 3.4334940189251917,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 14422
    },
    {
      "epoch": 3.4337320716538713,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 14423
    },
    {
      "epoch": 3.433970124382551,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 14424
    },
    {
      "epoch": 3.43420817711123,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3498,
      "step": 14425
    },
    {
      "epoch": 3.4344462298399097,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 14426
    },
    {
      "epoch": 3.434684282568589,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4246,
      "step": 14427
    },
    {
      "epoch": 3.4349223352972684,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.6122,
      "step": 14428
    },
    {
      "epoch": 3.4351603880259476,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2466,
      "step": 14429
    },
    {
      "epoch": 3.435398440754627,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 14430
    },
    {
      "epoch": 3.4356364934833064,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 14431
    },
    {
      "epoch": 3.435874546211986,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 14432
    },
    {
      "epoch": 3.4361125989406656,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 14433
    },
    {
      "epoch": 3.4363506516693447,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 14434
    },
    {
      "epoch": 3.4365887043980243,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3958,
      "step": 14435
    },
    {
      "epoch": 3.4368267571267035,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 14436
    },
    {
      "epoch": 3.437064809855383,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 14437
    },
    {
      "epoch": 3.4373028625840623,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5069,
      "step": 14438
    },
    {
      "epoch": 3.437540915312742,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 14439
    },
    {
      "epoch": 3.437778968041421,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 14440
    },
    {
      "epoch": 3.4380170207701006,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 14441
    },
    {
      "epoch": 3.43825507349878,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 14442
    },
    {
      "epoch": 3.4384931262274594,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 14443
    },
    {
      "epoch": 3.4387311789561386,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 14444
    },
    {
      "epoch": 3.438969231684818,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 14445
    },
    {
      "epoch": 3.4392072844134978,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 14446
    },
    {
      "epoch": 3.439445337142177,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 14447
    },
    {
      "epoch": 3.4396833898708565,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 14448
    },
    {
      "epoch": 3.4399214425995357,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 14449
    },
    {
      "epoch": 3.4401594953282153,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 14450
    },
    {
      "epoch": 3.4403975480568945,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 14451
    },
    {
      "epoch": 3.440635600785574,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 14452
    },
    {
      "epoch": 3.4408736535142532,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 14453
    },
    {
      "epoch": 3.441111706242933,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 14454
    },
    {
      "epoch": 3.4413497589716124,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 14455
    },
    {
      "epoch": 3.4415878117002916,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 14456
    },
    {
      "epoch": 3.441825864428971,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 14457
    },
    {
      "epoch": 3.4420639171576504,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.405,
      "step": 14458
    },
    {
      "epoch": 3.44230196988633,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 14459
    },
    {
      "epoch": 3.442540022615009,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 14460
    },
    {
      "epoch": 3.4427780753436887,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 14461
    },
    {
      "epoch": 3.443016128072368,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 14462
    },
    {
      "epoch": 3.4432541808010475,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 14463
    },
    {
      "epoch": 3.4434922335297267,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 14464
    },
    {
      "epoch": 3.4437302862584063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 14465
    },
    {
      "epoch": 3.4439683389870854,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 14466
    },
    {
      "epoch": 3.444206391715765,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 14467
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 14468
    },
    {
      "epoch": 3.444682497173124,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 14469
    },
    {
      "epoch": 3.4449205499018034,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 14470
    },
    {
      "epoch": 3.4451586026304826,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 14471
    },
    {
      "epoch": 3.445396655359162,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 14472
    },
    {
      "epoch": 3.4456347080878413,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 14473
    },
    {
      "epoch": 3.445872760816521,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2922,
      "step": 14474
    },
    {
      "epoch": 3.4461108135452,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 14475
    },
    {
      "epoch": 3.4463488662738797,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 14476
    },
    {
      "epoch": 3.4465869190025593,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 14477
    },
    {
      "epoch": 3.4468249717312385,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 14478
    },
    {
      "epoch": 3.4470630244599176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 14479
    },
    {
      "epoch": 3.4473010771885972,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 14480
    },
    {
      "epoch": 3.447539129917277,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 14481
    },
    {
      "epoch": 3.447777182645956,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 14482
    },
    {
      "epoch": 3.4480152353746356,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 14483
    },
    {
      "epoch": 3.4482532881033148,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 14484
    },
    {
      "epoch": 3.4484913408319944,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 14485
    },
    {
      "epoch": 3.4487293935606735,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 14486
    },
    {
      "epoch": 3.448967446289353,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 14487
    },
    {
      "epoch": 3.4492054990180323,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 14488
    },
    {
      "epoch": 3.449443551746712,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 14489
    },
    {
      "epoch": 3.4496816044753915,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 14490
    },
    {
      "epoch": 3.4499196572040707,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 14491
    },
    {
      "epoch": 3.4501577099327503,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 14492
    },
    {
      "epoch": 3.4503957626614294,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 14493
    },
    {
      "epoch": 3.450633815390109,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 14494
    },
    {
      "epoch": 3.450871868118788,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4569,
      "step": 14495
    },
    {
      "epoch": 3.451109920847468,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 14496
    },
    {
      "epoch": 3.451347973576147,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 14497
    },
    {
      "epoch": 3.4515860263048266,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3964,
      "step": 14498
    },
    {
      "epoch": 3.4518240790335057,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 14499
    },
    {
      "epoch": 3.4520621317621853,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 14500
    },
    {
      "epoch": 3.4520621317621853,
      "eval_runtime": 0.5366,
      "eval_samples_per_second": 1.863,
      "eval_steps_per_second": 1.863,
      "step": 14500
    },
    {
      "epoch": 3.4523001844908645,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 14501
    },
    {
      "epoch": 3.452538237219544,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 14502
    },
    {
      "epoch": 3.4527762899482237,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 14503
    },
    {
      "epoch": 3.453014342676903,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 14504
    },
    {
      "epoch": 3.4532523954055825,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 14505
    },
    {
      "epoch": 3.4534904481342616,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 14506
    },
    {
      "epoch": 3.4537285008629413,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 14507
    },
    {
      "epoch": 3.4539665535916204,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 14508
    },
    {
      "epoch": 3.4542046063203,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 14509
    },
    {
      "epoch": 3.454442659048979,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 14510
    },
    {
      "epoch": 3.454680711777659,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 14511
    },
    {
      "epoch": 3.4549187645063384,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3451,
      "step": 14512
    },
    {
      "epoch": 3.4551568172350176,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4595,
      "step": 14513
    },
    {
      "epoch": 3.4553948699636967,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 14514
    },
    {
      "epoch": 3.4556329226923763,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 14515
    },
    {
      "epoch": 3.455870975421056,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 14516
    },
    {
      "epoch": 3.456109028149735,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 14517
    },
    {
      "epoch": 3.4563470808784147,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 14518
    },
    {
      "epoch": 3.456585133607094,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 14519
    },
    {
      "epoch": 3.4568231863357735,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 14520
    },
    {
      "epoch": 3.4570612390644526,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 14521
    },
    {
      "epoch": 3.4572992917931322,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 14522
    },
    {
      "epoch": 3.4575373445218114,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 14523
    },
    {
      "epoch": 3.457775397250491,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 14524
    },
    {
      "epoch": 3.4580134499791706,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.361,
      "step": 14525
    },
    {
      "epoch": 3.4582515027078498,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 14526
    },
    {
      "epoch": 3.4584895554365294,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 14527
    },
    {
      "epoch": 3.4587276081652085,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 14528
    },
    {
      "epoch": 3.458965660893888,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2657,
      "step": 14529
    },
    {
      "epoch": 3.4592037136225673,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 14530
    },
    {
      "epoch": 3.459441766351247,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 14531
    },
    {
      "epoch": 3.459679819079926,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 14532
    },
    {
      "epoch": 3.4599178718086057,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 14533
    },
    {
      "epoch": 3.460155924537285,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 14534
    },
    {
      "epoch": 3.4603939772659644,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 14535
    },
    {
      "epoch": 3.4606320299946436,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 14536
    },
    {
      "epoch": 3.460870082723323,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 14537
    },
    {
      "epoch": 3.461108135452003,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 14538
    },
    {
      "epoch": 3.461346188180682,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 14539
    },
    {
      "epoch": 3.4615842409093616,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 14540
    },
    {
      "epoch": 3.4618222936380407,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 14541
    },
    {
      "epoch": 3.4620603463667203,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 14542
    },
    {
      "epoch": 3.4622983990953995,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 14543
    },
    {
      "epoch": 3.462536451824079,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 14544
    },
    {
      "epoch": 3.4627745045527583,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 14545
    },
    {
      "epoch": 3.463012557281438,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 14546
    },
    {
      "epoch": 3.4632506100101175,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 14547
    },
    {
      "epoch": 3.4634886627387966,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 14548
    },
    {
      "epoch": 3.4637267154674762,
      "grad_norm": 8.75,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 14549
    },
    {
      "epoch": 3.4639647681961554,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 14550
    },
    {
      "epoch": 3.464202820924835,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 14551
    },
    {
      "epoch": 3.464440873653514,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 14552
    },
    {
      "epoch": 3.4646789263821938,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 14553
    },
    {
      "epoch": 3.464916979110873,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 14554
    },
    {
      "epoch": 3.4651550318395525,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 14555
    },
    {
      "epoch": 3.4653930845682317,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 14556
    },
    {
      "epoch": 3.4656311372969113,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4771,
      "step": 14557
    },
    {
      "epoch": 3.4658691900255905,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 14558
    },
    {
      "epoch": 3.46610724275427,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 14559
    },
    {
      "epoch": 3.4663452954829497,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 14560
    },
    {
      "epoch": 3.466583348211629,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 14561
    },
    {
      "epoch": 3.4668214009403084,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 14562
    },
    {
      "epoch": 3.4670594536689876,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 14563
    },
    {
      "epoch": 3.467297506397667,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 14564
    },
    {
      "epoch": 3.4675355591263464,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3486,
      "step": 14565
    },
    {
      "epoch": 3.467773611855026,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 14566
    },
    {
      "epoch": 3.468011664583705,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 14567
    },
    {
      "epoch": 3.4682497173123847,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.68,
      "step": 14568
    },
    {
      "epoch": 3.4684877700410643,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5599,
      "step": 14569
    },
    {
      "epoch": 3.4687258227697435,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 14570
    },
    {
      "epoch": 3.4689638754984227,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 14571
    },
    {
      "epoch": 3.4692019282271023,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 14572
    },
    {
      "epoch": 3.469439980955782,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 14573
    },
    {
      "epoch": 3.469678033684461,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.428,
      "step": 14574
    },
    {
      "epoch": 3.4699160864131406,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 14575
    },
    {
      "epoch": 3.47015413914182,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 14576
    },
    {
      "epoch": 3.4703921918704994,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 14577
    },
    {
      "epoch": 3.4706302445991786,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 14578
    },
    {
      "epoch": 3.470868297327858,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 14579
    },
    {
      "epoch": 3.4711063500565373,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 14580
    },
    {
      "epoch": 3.471344402785217,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 14581
    },
    {
      "epoch": 3.4715824555138965,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 14582
    },
    {
      "epoch": 3.4718205082425757,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 14583
    },
    {
      "epoch": 3.4720585609712553,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 14584
    },
    {
      "epoch": 3.4722966136999345,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 14585
    },
    {
      "epoch": 3.472534666428614,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 14586
    },
    {
      "epoch": 3.4727727191572932,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2913,
      "step": 14587
    },
    {
      "epoch": 3.473010771885973,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 14588
    },
    {
      "epoch": 3.473248824614652,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 14589
    },
    {
      "epoch": 3.4734868773433316,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 14590
    },
    {
      "epoch": 3.4737249300720108,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.444,
      "step": 14591
    },
    {
      "epoch": 3.4739629828006904,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 14592
    },
    {
      "epoch": 3.4742010355293695,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 14593
    },
    {
      "epoch": 3.474439088258049,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3149,
      "step": 14594
    },
    {
      "epoch": 3.4746771409867288,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 14595
    },
    {
      "epoch": 3.474915193715408,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 14596
    },
    {
      "epoch": 3.4751532464440875,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 14597
    },
    {
      "epoch": 3.4753912991727667,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 14598
    },
    {
      "epoch": 3.4756293519014463,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 14599
    },
    {
      "epoch": 3.4758674046301254,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 14600
    },
    {
      "epoch": 3.4758674046301254,
      "eval_runtime": 0.523,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 14600
    },
    {
      "epoch": 3.476105457358805,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 14601
    },
    {
      "epoch": 3.476343510087484,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 14602
    },
    {
      "epoch": 3.476581562816164,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 14603
    },
    {
      "epoch": 3.4768196155448434,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.701,
      "step": 14604
    },
    {
      "epoch": 3.4770576682735226,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 14605
    },
    {
      "epoch": 3.4772957210022017,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.6869,
      "step": 14606
    },
    {
      "epoch": 3.4775337737308814,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 14607
    },
    {
      "epoch": 3.477771826459561,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 14608
    },
    {
      "epoch": 3.47800987918824,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 14609
    },
    {
      "epoch": 3.4782479319169197,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 14610
    },
    {
      "epoch": 3.478485984645599,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 14611
    },
    {
      "epoch": 3.4787240373742785,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 14612
    },
    {
      "epoch": 3.4789620901029576,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 14613
    },
    {
      "epoch": 3.4792001428316373,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 14614
    },
    {
      "epoch": 3.4794381955603164,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 14615
    },
    {
      "epoch": 3.479676248288996,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 14616
    },
    {
      "epoch": 3.4799143010176756,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.5312,
      "step": 14617
    },
    {
      "epoch": 3.480152353746355,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 14618
    },
    {
      "epoch": 3.4803904064750344,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 14619
    },
    {
      "epoch": 3.4806284592037136,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 14620
    },
    {
      "epoch": 3.480866511932393,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 14621
    },
    {
      "epoch": 3.4811045646610723,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 14622
    },
    {
      "epoch": 3.481342617389752,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 14623
    },
    {
      "epoch": 3.481580670118431,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7516,
      "step": 14624
    },
    {
      "epoch": 3.4818187228471107,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 14625
    },
    {
      "epoch": 3.48205677557579,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 14626
    },
    {
      "epoch": 3.4822948283044695,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 14627
    },
    {
      "epoch": 3.4825328810331486,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 14628
    },
    {
      "epoch": 3.4827709337618282,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 14629
    },
    {
      "epoch": 3.483008986490508,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 14630
    },
    {
      "epoch": 3.483247039219187,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 14631
    },
    {
      "epoch": 3.4834850919478666,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 14632
    },
    {
      "epoch": 3.4837231446765458,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 14633
    },
    {
      "epoch": 3.4839611974052254,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 14634
    },
    {
      "epoch": 3.4841992501339045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 14635
    },
    {
      "epoch": 3.484437302862584,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 14636
    },
    {
      "epoch": 3.4846753555912633,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 14637
    },
    {
      "epoch": 3.484913408319943,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 14638
    },
    {
      "epoch": 3.4851514610486225,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 14639
    },
    {
      "epoch": 3.4853895137773017,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 14640
    },
    {
      "epoch": 3.4856275665059813,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 14641
    },
    {
      "epoch": 3.4858656192346604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 14642
    },
    {
      "epoch": 3.48610367196334,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 14643
    },
    {
      "epoch": 3.486341724692019,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 14644
    },
    {
      "epoch": 3.486579777420699,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 14645
    },
    {
      "epoch": 3.486817830149378,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 14646
    },
    {
      "epoch": 3.4870558828780576,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 14647
    },
    {
      "epoch": 3.4872939356067367,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 14648
    },
    {
      "epoch": 3.4875319883354163,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 14649
    },
    {
      "epoch": 3.4877700410640955,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8676,
      "step": 14650
    },
    {
      "epoch": 3.488008093792775,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 14651
    },
    {
      "epoch": 3.4882461465214547,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 14652
    },
    {
      "epoch": 3.488484199250134,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 14653
    },
    {
      "epoch": 3.4887222519788135,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 14654
    },
    {
      "epoch": 3.4889603047074926,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 14655
    },
    {
      "epoch": 3.4891983574361722,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 14656
    },
    {
      "epoch": 3.4894364101648514,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 14657
    },
    {
      "epoch": 3.489674462893531,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 14658
    },
    {
      "epoch": 3.48991251562221,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 14659
    },
    {
      "epoch": 3.4901505683508898,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 14660
    },
    {
      "epoch": 3.490388621079569,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4299,
      "step": 14661
    },
    {
      "epoch": 3.4906266738082485,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 14662
    },
    {
      "epoch": 3.4908647265369277,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 14663
    },
    {
      "epoch": 3.4911027792656073,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3403,
      "step": 14664
    },
    {
      "epoch": 3.491340831994287,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 14665
    },
    {
      "epoch": 3.491578884722966,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 14666
    },
    {
      "epoch": 3.4918169374516457,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 14667
    },
    {
      "epoch": 3.492054990180325,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 14668
    },
    {
      "epoch": 3.4922930429090044,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 14669
    },
    {
      "epoch": 3.4925310956376836,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 14670
    },
    {
      "epoch": 3.492769148366363,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 14671
    },
    {
      "epoch": 3.4930072010950424,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 14672
    },
    {
      "epoch": 3.493245253823722,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 14673
    },
    {
      "epoch": 3.4934833065524016,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 14674
    },
    {
      "epoch": 3.4937213592810807,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 14675
    },
    {
      "epoch": 3.4939594120097603,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 14676
    },
    {
      "epoch": 3.4941974647384395,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3878,
      "step": 14677
    },
    {
      "epoch": 3.494435517467119,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 14678
    },
    {
      "epoch": 3.4946735701957983,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 14679
    },
    {
      "epoch": 3.494911622924478,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 14680
    },
    {
      "epoch": 3.495149675653157,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 14681
    },
    {
      "epoch": 3.4953877283818366,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 14682
    },
    {
      "epoch": 3.495625781110516,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3859,
      "step": 14683
    },
    {
      "epoch": 3.4958638338391954,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 14684
    },
    {
      "epoch": 3.4961018865678746,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 14685
    },
    {
      "epoch": 3.496339939296554,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 14686
    },
    {
      "epoch": 3.496577992025234,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 14687
    },
    {
      "epoch": 3.496816044753913,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.272,
      "step": 14688
    },
    {
      "epoch": 3.4970540974825925,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 14689
    },
    {
      "epoch": 3.4972921502112717,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 14690
    },
    {
      "epoch": 3.4975302029399513,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 14691
    },
    {
      "epoch": 3.4977682556686305,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 14692
    },
    {
      "epoch": 3.49800630839731,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 14693
    },
    {
      "epoch": 3.4982443611259892,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8408,
      "step": 14694
    },
    {
      "epoch": 3.498482413854669,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 14695
    },
    {
      "epoch": 3.4987204665833485,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 14696
    },
    {
      "epoch": 3.4989585193120276,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 14697
    },
    {
      "epoch": 3.4991965720407068,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 14698
    },
    {
      "epoch": 3.4994346247693864,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 14699
    },
    {
      "epoch": 3.499672677498066,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 14700
    },
    {
      "epoch": 3.499672677498066,
      "eval_runtime": 0.5135,
      "eval_samples_per_second": 1.947,
      "eval_steps_per_second": 1.947,
      "step": 14700
    },
    {
      "epoch": 3.499910730226745,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 14701
    },
    {
      "epoch": 3.5001487829554248,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 14702
    },
    {
      "epoch": 3.500386835684104,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3634,
      "step": 14703
    },
    {
      "epoch": 3.5006248884127835,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 14704
    },
    {
      "epoch": 3.5008629411414627,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 14705
    },
    {
      "epoch": 3.5011009938701423,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 14706
    },
    {
      "epoch": 3.5013390465988214,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 14707
    },
    {
      "epoch": 3.501577099327501,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 14708
    },
    {
      "epoch": 3.5018151520561807,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 14709
    },
    {
      "epoch": 3.50205320478486,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 14710
    },
    {
      "epoch": 3.502291257513539,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 14711
    },
    {
      "epoch": 3.5025293102422186,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2693,
      "step": 14712
    },
    {
      "epoch": 3.502767362970898,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4882,
      "step": 14713
    },
    {
      "epoch": 3.5030054156995774,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 14714
    },
    {
      "epoch": 3.503243468428257,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2933,
      "step": 14715
    },
    {
      "epoch": 3.503481521156936,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3859,
      "step": 14716
    },
    {
      "epoch": 3.5037195738856157,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 14717
    },
    {
      "epoch": 3.5039576266142953,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 14718
    },
    {
      "epoch": 3.5041956793429745,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2542,
      "step": 14719
    },
    {
      "epoch": 3.5044337320716537,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 14720
    },
    {
      "epoch": 3.5046717848003333,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 14721
    },
    {
      "epoch": 3.504909837529013,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 14722
    },
    {
      "epoch": 3.505147890257692,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 14723
    },
    {
      "epoch": 3.5053859429863716,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 14724
    },
    {
      "epoch": 3.505623995715051,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 14725
    },
    {
      "epoch": 3.5058620484437304,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 14726
    },
    {
      "epoch": 3.5061001011724096,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 14727
    },
    {
      "epoch": 3.506338153901089,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 14728
    },
    {
      "epoch": 3.5065762066297683,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 14729
    },
    {
      "epoch": 3.506814259358448,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 14730
    },
    {
      "epoch": 3.5070523120871275,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 14731
    },
    {
      "epoch": 3.5072903648158067,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 14732
    },
    {
      "epoch": 3.507528417544486,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 14733
    },
    {
      "epoch": 3.5077664702731655,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 14734
    },
    {
      "epoch": 3.508004523001845,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 14735
    },
    {
      "epoch": 3.5082425757305242,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.5268,
      "step": 14736
    },
    {
      "epoch": 3.508480628459204,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 14737
    },
    {
      "epoch": 3.508718681187883,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 14738
    },
    {
      "epoch": 3.5089567339165626,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 14739
    },
    {
      "epoch": 3.5091947866452418,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 14740
    },
    {
      "epoch": 3.5094328393739214,
      "grad_norm": 6.96875,
      "learning_rate": 3e-05,
      "loss": 2.6516,
      "step": 14741
    },
    {
      "epoch": 3.5096708921026005,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 14742
    },
    {
      "epoch": 3.50990894483128,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 14743
    },
    {
      "epoch": 3.5101469975599597,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 14744
    },
    {
      "epoch": 3.510385050288639,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 14745
    },
    {
      "epoch": 3.510623103017318,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 14746
    },
    {
      "epoch": 3.5108611557459977,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 14747
    },
    {
      "epoch": 3.5110992084746773,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 14748
    },
    {
      "epoch": 3.5113372612033564,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 14749
    },
    {
      "epoch": 3.511575313932036,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 14750
    },
    {
      "epoch": 3.511813366660715,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 14751
    },
    {
      "epoch": 3.512051419389395,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 14752
    },
    {
      "epoch": 3.5122894721180744,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 14753
    },
    {
      "epoch": 3.5125275248467536,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 14754
    },
    {
      "epoch": 3.5127655775754327,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 14755
    },
    {
      "epoch": 3.5130036303041123,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 14756
    },
    {
      "epoch": 3.513241683032792,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 14757
    },
    {
      "epoch": 3.513479735761471,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 14758
    },
    {
      "epoch": 3.5137177884901507,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 14759
    },
    {
      "epoch": 3.51395584121883,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 14760
    },
    {
      "epoch": 3.5141938939475095,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 14761
    },
    {
      "epoch": 3.5144319466761886,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 14762
    },
    {
      "epoch": 3.5146699994048682,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3491,
      "step": 14763
    },
    {
      "epoch": 3.5149080521335474,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 14764
    },
    {
      "epoch": 3.515146104862227,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 14765
    },
    {
      "epoch": 3.5153841575909066,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 14766
    },
    {
      "epoch": 3.5156222103195858,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 14767
    },
    {
      "epoch": 3.515860263048265,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 14768
    },
    {
      "epoch": 3.5160983157769445,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 14769
    },
    {
      "epoch": 3.516336368505624,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3907,
      "step": 14770
    },
    {
      "epoch": 3.5165744212343033,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 14771
    },
    {
      "epoch": 3.516812473962983,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 14772
    },
    {
      "epoch": 3.517050526691662,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 14773
    },
    {
      "epoch": 3.5172885794203417,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 14774
    },
    {
      "epoch": 3.5175266321490213,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 14775
    },
    {
      "epoch": 3.5177646848777004,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 14776
    },
    {
      "epoch": 3.5180027376063796,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 14777
    },
    {
      "epoch": 3.518240790335059,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 14778
    },
    {
      "epoch": 3.518478843063739,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 14779
    },
    {
      "epoch": 3.518716895792418,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 14780
    },
    {
      "epoch": 3.5189549485210976,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 14781
    },
    {
      "epoch": 3.5191930012497767,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 14782
    },
    {
      "epoch": 3.5194310539784563,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 14783
    },
    {
      "epoch": 3.5196691067071355,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2737,
      "step": 14784
    },
    {
      "epoch": 3.519907159435815,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 14785
    },
    {
      "epoch": 3.5201452121644943,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 14786
    },
    {
      "epoch": 3.520383264893174,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 14787
    },
    {
      "epoch": 3.5206213176218535,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 14788
    },
    {
      "epoch": 3.5208593703505326,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 14789
    },
    {
      "epoch": 3.521097423079212,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 14790
    },
    {
      "epoch": 3.5213354758078914,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 14791
    },
    {
      "epoch": 3.521573528536571,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 14792
    },
    {
      "epoch": 3.52181158126525,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 14793
    },
    {
      "epoch": 3.52204963399393,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 14794
    },
    {
      "epoch": 3.522287686722609,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 14795
    },
    {
      "epoch": 3.5225257394512886,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 14796
    },
    {
      "epoch": 3.5227637921799677,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 14797
    },
    {
      "epoch": 3.5230018449086473,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 14798
    },
    {
      "epoch": 3.5232398976373265,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 14799
    },
    {
      "epoch": 3.523477950366006,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 14800
    },
    {
      "epoch": 3.523477950366006,
      "eval_runtime": 0.5375,
      "eval_samples_per_second": 1.86,
      "eval_steps_per_second": 1.86,
      "step": 14800
    },
    {
      "epoch": 3.5237160030946857,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3176,
      "step": 14801
    },
    {
      "epoch": 3.523954055823365,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 14802
    },
    {
      "epoch": 3.524192108552044,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 14803
    },
    {
      "epoch": 3.5244301612807236,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 14804
    },
    {
      "epoch": 3.5246682140094032,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 14805
    },
    {
      "epoch": 3.5249062667380824,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 14806
    },
    {
      "epoch": 3.525144319466762,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 14807
    },
    {
      "epoch": 3.525382372195441,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 14808
    },
    {
      "epoch": 3.5256204249241208,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 14809
    },
    {
      "epoch": 3.5258584776528004,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 14810
    },
    {
      "epoch": 3.5260965303814795,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 14811
    },
    {
      "epoch": 3.5263345831101587,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 14812
    },
    {
      "epoch": 3.5265726358388383,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 14813
    },
    {
      "epoch": 3.526810688567518,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 14814
    },
    {
      "epoch": 3.527048741296197,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 14815
    },
    {
      "epoch": 3.5272867940248767,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 14816
    },
    {
      "epoch": 3.527524846753556,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 14817
    },
    {
      "epoch": 3.5277628994822354,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 14818
    },
    {
      "epoch": 3.5280009522109146,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 14819
    },
    {
      "epoch": 3.528239004939594,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 14820
    },
    {
      "epoch": 3.5284770576682734,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 14821
    },
    {
      "epoch": 3.528715110396953,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7558,
      "step": 14822
    },
    {
      "epoch": 3.5289531631256326,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 14823
    },
    {
      "epoch": 3.5291912158543117,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 14824
    },
    {
      "epoch": 3.529429268582991,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 14825
    },
    {
      "epoch": 3.5296673213116705,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 14826
    },
    {
      "epoch": 3.52990537404035,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 14827
    },
    {
      "epoch": 3.5301434267690293,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 14828
    },
    {
      "epoch": 3.530381479497709,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 14829
    },
    {
      "epoch": 3.530619532226388,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 14830
    },
    {
      "epoch": 3.5308575849550676,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 14831
    },
    {
      "epoch": 3.531095637683747,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 14832
    },
    {
      "epoch": 3.5313336904124264,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 14833
    },
    {
      "epoch": 3.5315717431411056,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 14834
    },
    {
      "epoch": 3.531809795869785,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 14835
    },
    {
      "epoch": 3.5320478485984648,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2846,
      "step": 14836
    },
    {
      "epoch": 3.532285901327144,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 14837
    },
    {
      "epoch": 3.532523954055823,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 14838
    },
    {
      "epoch": 3.5327620067845027,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 14839
    },
    {
      "epoch": 3.5330000595131823,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 14840
    },
    {
      "epoch": 3.5332381122418615,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 14841
    },
    {
      "epoch": 3.533476164970541,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 14842
    },
    {
      "epoch": 3.5337142176992202,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 14843
    },
    {
      "epoch": 3.5339522704279,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 14844
    },
    {
      "epoch": 3.5341903231565794,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 14845
    },
    {
      "epoch": 3.5344283758852586,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 14846
    },
    {
      "epoch": 3.5346664286139378,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 14847
    },
    {
      "epoch": 3.5349044813426174,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 14848
    },
    {
      "epoch": 3.535142534071297,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 14849
    },
    {
      "epoch": 3.535380586799976,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 14850
    },
    {
      "epoch": 3.5356186395286557,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 14851
    },
    {
      "epoch": 3.535856692257335,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 14852
    },
    {
      "epoch": 3.5360947449860145,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 14853
    },
    {
      "epoch": 3.5363327977146937,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 14854
    },
    {
      "epoch": 3.5365708504433733,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 14855
    },
    {
      "epoch": 3.5368089031720524,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2804,
      "step": 14856
    },
    {
      "epoch": 3.537046955900732,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3485,
      "step": 14857
    },
    {
      "epoch": 3.5372850086294116,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3631,
      "step": 14858
    },
    {
      "epoch": 3.537523061358091,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 14859
    },
    {
      "epoch": 3.53776111408677,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4495,
      "step": 14860
    },
    {
      "epoch": 3.5379991668154496,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 14861
    },
    {
      "epoch": 3.538237219544129,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 14862
    },
    {
      "epoch": 3.5384752722728083,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 14863
    },
    {
      "epoch": 3.538713325001488,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 14864
    },
    {
      "epoch": 3.538951377730167,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 14865
    },
    {
      "epoch": 3.5391894304588467,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 14866
    },
    {
      "epoch": 3.5394274831875263,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 14867
    },
    {
      "epoch": 3.5396655359162055,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 14868
    },
    {
      "epoch": 3.5399035886448846,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 14869
    },
    {
      "epoch": 3.5401416413735642,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3242,
      "step": 14870
    },
    {
      "epoch": 3.540379694102244,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8096,
      "step": 14871
    },
    {
      "epoch": 3.540617746830923,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5005,
      "step": 14872
    },
    {
      "epoch": 3.5408557995596026,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3033,
      "step": 14873
    },
    {
      "epoch": 3.5410938522882818,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 14874
    },
    {
      "epoch": 3.5413319050169614,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 14875
    },
    {
      "epoch": 3.5415699577456405,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 14876
    },
    {
      "epoch": 3.54180801047432,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 14877
    },
    {
      "epoch": 3.5420460632029993,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 14878
    },
    {
      "epoch": 3.542284115931679,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 14879
    },
    {
      "epoch": 3.5425221686603585,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 14880
    },
    {
      "epoch": 3.5427602213890377,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 14881
    },
    {
      "epoch": 3.542998274117717,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 14882
    },
    {
      "epoch": 3.5432363268463964,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 14883
    },
    {
      "epoch": 3.543474379575076,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 14884
    },
    {
      "epoch": 3.543712432303755,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 14885
    },
    {
      "epoch": 3.543950485032435,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 14886
    },
    {
      "epoch": 3.544188537761114,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 14887
    },
    {
      "epoch": 3.5444265904897936,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 14888
    },
    {
      "epoch": 3.5446646432184727,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 14889
    },
    {
      "epoch": 3.5449026959471523,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3033,
      "step": 14890
    },
    {
      "epoch": 3.5451407486758315,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 14891
    },
    {
      "epoch": 3.545378801404511,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3094,
      "step": 14892
    },
    {
      "epoch": 3.5456168541331907,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 14893
    },
    {
      "epoch": 3.54585490686187,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3176,
      "step": 14894
    },
    {
      "epoch": 3.546092959590549,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 14895
    },
    {
      "epoch": 3.5463310123192286,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 14896
    },
    {
      "epoch": 3.5465690650479083,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3904,
      "step": 14897
    },
    {
      "epoch": 3.5468071177765874,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 14898
    },
    {
      "epoch": 3.547045170505267,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.491,
      "step": 14899
    },
    {
      "epoch": 3.547283223233946,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.364,
      "step": 14900
    },
    {
      "epoch": 3.547283223233946,
      "eval_runtime": 0.5342,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 1.872,
      "step": 14900
    },
    {
      "epoch": 3.547521275962626,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 14901
    },
    {
      "epoch": 3.5477593286913054,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 14902
    },
    {
      "epoch": 3.5479973814199846,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 14903
    },
    {
      "epoch": 3.5482354341486637,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3271,
      "step": 14904
    },
    {
      "epoch": 3.5484734868773433,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 14905
    },
    {
      "epoch": 3.548711539606023,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 14906
    },
    {
      "epoch": 3.548949592334702,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 14907
    },
    {
      "epoch": 3.5491876450633817,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 14908
    },
    {
      "epoch": 3.549425697792061,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 14909
    },
    {
      "epoch": 3.5496637505207405,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 14910
    },
    {
      "epoch": 3.5499018032494196,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.341,
      "step": 14911
    },
    {
      "epoch": 3.5501398559780992,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 14912
    },
    {
      "epoch": 3.5503779087067784,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 14913
    },
    {
      "epoch": 3.550615961435458,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 14914
    },
    {
      "epoch": 3.5508540141641376,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 14915
    },
    {
      "epoch": 3.5510920668928168,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 14916
    },
    {
      "epoch": 3.551330119621496,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 14917
    },
    {
      "epoch": 3.5515681723501755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 14918
    },
    {
      "epoch": 3.551806225078855,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 14919
    },
    {
      "epoch": 3.5520442778075343,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 14920
    },
    {
      "epoch": 3.552282330536214,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 14921
    },
    {
      "epoch": 3.552520383264893,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 14922
    },
    {
      "epoch": 3.5527584359935727,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 14923
    },
    {
      "epoch": 3.552996488722252,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 14924
    },
    {
      "epoch": 3.5532345414509314,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 14925
    },
    {
      "epoch": 3.5534725941796106,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 14926
    },
    {
      "epoch": 3.55371064690829,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 14927
    },
    {
      "epoch": 3.55394869963697,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 14928
    },
    {
      "epoch": 3.554186752365649,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 14929
    },
    {
      "epoch": 3.554424805094328,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 14930
    },
    {
      "epoch": 3.5546628578230077,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3867,
      "step": 14931
    },
    {
      "epoch": 3.5549009105516873,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3776,
      "step": 14932
    },
    {
      "epoch": 3.5551389632803665,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 14933
    },
    {
      "epoch": 3.555377016009046,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 14934
    },
    {
      "epoch": 3.5556150687377253,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 14935
    },
    {
      "epoch": 3.555853121466405,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 14936
    },
    {
      "epoch": 3.5560911741950845,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 14937
    },
    {
      "epoch": 3.5563292269237636,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 14938
    },
    {
      "epoch": 3.556567279652443,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 14939
    },
    {
      "epoch": 3.5568053323811224,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 14940
    },
    {
      "epoch": 3.557043385109802,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 14941
    },
    {
      "epoch": 3.557281437838481,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 14942
    },
    {
      "epoch": 3.5575194905671608,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 14943
    },
    {
      "epoch": 3.55775754329584,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4447,
      "step": 14944
    },
    {
      "epoch": 3.5579955960245195,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 14945
    },
    {
      "epoch": 3.5582336487531987,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 14946
    },
    {
      "epoch": 3.5584717014818783,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 14947
    },
    {
      "epoch": 3.5587097542105575,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 14948
    },
    {
      "epoch": 3.558947806939237,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 14949
    },
    {
      "epoch": 3.5591858596679167,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 14950
    },
    {
      "epoch": 3.559423912396596,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4077,
      "step": 14951
    },
    {
      "epoch": 3.559661965125275,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3604,
      "step": 14952
    },
    {
      "epoch": 3.5599000178539546,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 14953
    },
    {
      "epoch": 3.560138070582634,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3845,
      "step": 14954
    },
    {
      "epoch": 3.5603761233113134,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 14955
    },
    {
      "epoch": 3.560614176039993,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 14956
    },
    {
      "epoch": 3.560852228768672,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 14957
    },
    {
      "epoch": 3.5610902814973517,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 14958
    },
    {
      "epoch": 3.561328334226031,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 14959
    },
    {
      "epoch": 3.5615663869547105,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 14960
    },
    {
      "epoch": 3.5618044396833897,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 14961
    },
    {
      "epoch": 3.5620424924120693,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 14962
    },
    {
      "epoch": 3.562280545140749,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 14963
    },
    {
      "epoch": 3.562518597869428,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 14964
    },
    {
      "epoch": 3.5627566505981076,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 14965
    },
    {
      "epoch": 3.562994703326787,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 14966
    },
    {
      "epoch": 3.5632327560554664,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 14967
    },
    {
      "epoch": 3.5634708087841456,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 14968
    },
    {
      "epoch": 3.563708861512825,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 14969
    },
    {
      "epoch": 3.5639469142415043,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 14970
    },
    {
      "epoch": 3.564184966970184,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 14971
    },
    {
      "epoch": 3.5644230196988635,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 14972
    },
    {
      "epoch": 3.5646610724275427,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3204,
      "step": 14973
    },
    {
      "epoch": 3.564899125156222,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.429,
      "step": 14974
    },
    {
      "epoch": 3.5651371778849015,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 14975
    },
    {
      "epoch": 3.565375230613581,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 14976
    },
    {
      "epoch": 3.5656132833422602,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 14977
    },
    {
      "epoch": 3.56585133607094,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 14978
    },
    {
      "epoch": 3.566089388799619,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 14979
    },
    {
      "epoch": 3.5663274415282986,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 14980
    },
    {
      "epoch": 3.5665654942569778,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 14981
    },
    {
      "epoch": 3.5668035469856574,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 14982
    },
    {
      "epoch": 3.5670415997143365,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 14983
    },
    {
      "epoch": 3.567279652443016,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 14984
    },
    {
      "epoch": 3.5675177051716958,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 14985
    },
    {
      "epoch": 3.567755757900375,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 14986
    },
    {
      "epoch": 3.567993810629054,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 14987
    },
    {
      "epoch": 3.5682318633577337,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 14988
    },
    {
      "epoch": 3.5684699160864133,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.7256,
      "step": 14989
    },
    {
      "epoch": 3.5687079688150924,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 14990
    },
    {
      "epoch": 3.568946021543772,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 14991
    },
    {
      "epoch": 3.569184074272451,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 14992
    },
    {
      "epoch": 3.569422127001131,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 14993
    },
    {
      "epoch": 3.5696601797298104,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 14994
    },
    {
      "epoch": 3.5698982324584896,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 14995
    },
    {
      "epoch": 3.5701362851871687,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 14996
    },
    {
      "epoch": 3.5703743379158484,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 14997
    },
    {
      "epoch": 3.570612390644528,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2893,
      "step": 14998
    },
    {
      "epoch": 3.570850443373207,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 14999
    },
    {
      "epoch": 3.5710884961018867,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 15000
    },
    {
      "epoch": 3.5710884961018867,
      "eval_runtime": 0.5212,
      "eval_samples_per_second": 1.919,
      "eval_steps_per_second": 1.919,
      "step": 15000
    },
    {
      "epoch": 3.571326548830566,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4998,
      "step": 15001
    },
    {
      "epoch": 3.5715646015592455,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 15002
    },
    {
      "epoch": 3.5718026542879246,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 15003
    },
    {
      "epoch": 3.5720407070166043,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 15004
    },
    {
      "epoch": 3.5722787597452834,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 15005
    },
    {
      "epoch": 3.572516812473963,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 15006
    },
    {
      "epoch": 3.5727548652026426,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2948,
      "step": 15007
    },
    {
      "epoch": 3.572992917931322,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 15008
    },
    {
      "epoch": 3.573230970660001,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 15009
    },
    {
      "epoch": 3.5734690233886806,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 15010
    },
    {
      "epoch": 3.57370707611736,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3154,
      "step": 15011
    },
    {
      "epoch": 3.5739451288460393,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4969,
      "step": 15012
    },
    {
      "epoch": 3.574183181574719,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3579,
      "step": 15013
    },
    {
      "epoch": 3.574421234303398,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 15014
    },
    {
      "epoch": 3.5746592870320777,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 15015
    },
    {
      "epoch": 3.574897339760757,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.7926,
      "step": 15016
    },
    {
      "epoch": 3.5751353924894365,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 15017
    },
    {
      "epoch": 3.5753734452181156,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 15018
    },
    {
      "epoch": 3.5756114979467952,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 15019
    },
    {
      "epoch": 3.575849550675475,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3501,
      "step": 15020
    },
    {
      "epoch": 3.576087603404154,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 15021
    },
    {
      "epoch": 3.576325656132833,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 15022
    },
    {
      "epoch": 3.5765637088615128,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 15023
    },
    {
      "epoch": 3.5768017615901924,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4301,
      "step": 15024
    },
    {
      "epoch": 3.5770398143188715,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 15025
    },
    {
      "epoch": 3.577277867047551,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 15026
    },
    {
      "epoch": 3.5775159197762303,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 15027
    },
    {
      "epoch": 3.57775397250491,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 15028
    },
    {
      "epoch": 3.5779920252335895,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 15029
    },
    {
      "epoch": 3.5782300779622687,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 15030
    },
    {
      "epoch": 3.578468130690948,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3852,
      "step": 15031
    },
    {
      "epoch": 3.5787061834196274,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 15032
    },
    {
      "epoch": 3.578944236148307,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 15033
    },
    {
      "epoch": 3.579182288876986,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3972,
      "step": 15034
    },
    {
      "epoch": 3.579420341605666,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 15035
    },
    {
      "epoch": 3.579658394334345,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4488,
      "step": 15036
    },
    {
      "epoch": 3.5798964470630246,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 15037
    },
    {
      "epoch": 3.5801344997917037,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 15038
    },
    {
      "epoch": 3.5803725525203833,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 15039
    },
    {
      "epoch": 3.5806106052490625,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 15040
    },
    {
      "epoch": 3.580848657977742,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 15041
    },
    {
      "epoch": 3.5810867107064217,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.324,
      "step": 15042
    },
    {
      "epoch": 3.581324763435101,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 15043
    },
    {
      "epoch": 3.58156281616378,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 15044
    },
    {
      "epoch": 3.5818008688924596,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 15045
    },
    {
      "epoch": 3.5820389216211392,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 15046
    },
    {
      "epoch": 3.5822769743498184,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 15047
    },
    {
      "epoch": 3.582515027078498,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 15048
    },
    {
      "epoch": 3.582753079807177,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 15049
    },
    {
      "epoch": 3.5829911325358568,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 15050
    },
    {
      "epoch": 3.583229185264536,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 15051
    },
    {
      "epoch": 3.5834672379932155,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 15052
    },
    {
      "epoch": 3.5837052907218947,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 15053
    },
    {
      "epoch": 3.5839433434505743,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 15054
    },
    {
      "epoch": 3.584181396179254,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 15055
    },
    {
      "epoch": 3.584419448907933,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 15056
    },
    {
      "epoch": 3.5846575016366127,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 15057
    },
    {
      "epoch": 3.584895554365292,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 15058
    },
    {
      "epoch": 3.5851336070939714,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 15059
    },
    {
      "epoch": 3.5853716598226506,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 15060
    },
    {
      "epoch": 3.58560971255133,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 15061
    },
    {
      "epoch": 3.5858477652800094,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 15062
    },
    {
      "epoch": 3.586085818008689,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 15063
    },
    {
      "epoch": 3.5863238707373686,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 15064
    },
    {
      "epoch": 3.5865619234660477,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 15065
    },
    {
      "epoch": 3.586799976194727,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 15066
    },
    {
      "epoch": 3.5870380289234065,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 15067
    },
    {
      "epoch": 3.587276081652086,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3215,
      "step": 15068
    },
    {
      "epoch": 3.5875141343807653,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 15069
    },
    {
      "epoch": 3.587752187109445,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.334,
      "step": 15070
    },
    {
      "epoch": 3.587990239838124,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 15071
    },
    {
      "epoch": 3.5882282925668036,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 15072
    },
    {
      "epoch": 3.588466345295483,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 15073
    },
    {
      "epoch": 3.5887043980241624,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 15074
    },
    {
      "epoch": 3.5889424507528416,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.5898,
      "step": 15075
    },
    {
      "epoch": 3.589180503481521,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 15076
    },
    {
      "epoch": 3.589418556210201,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 15077
    },
    {
      "epoch": 3.58965660893888,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 15078
    },
    {
      "epoch": 3.589894661667559,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 15079
    },
    {
      "epoch": 3.5901327143962387,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 15080
    },
    {
      "epoch": 3.5903707671249183,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3151,
      "step": 15081
    },
    {
      "epoch": 3.5906088198535975,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 15082
    },
    {
      "epoch": 3.590846872582277,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 15083
    },
    {
      "epoch": 3.5910849253109562,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 15084
    },
    {
      "epoch": 3.591322978039636,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 15085
    },
    {
      "epoch": 3.5915610307683155,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 15086
    },
    {
      "epoch": 3.5917990834969946,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 15087
    },
    {
      "epoch": 3.5920371362256738,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 15088
    },
    {
      "epoch": 3.5922751889543534,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 15089
    },
    {
      "epoch": 3.592513241683033,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 15090
    },
    {
      "epoch": 3.592751294411712,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3626,
      "step": 15091
    },
    {
      "epoch": 3.5929893471403918,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 15092
    },
    {
      "epoch": 3.593227399869071,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 15093
    },
    {
      "epoch": 3.5934654525977505,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 15094
    },
    {
      "epoch": 3.5937035053264297,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 15095
    },
    {
      "epoch": 3.5939415580551093,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 15096
    },
    {
      "epoch": 3.5941796107837884,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 15097
    },
    {
      "epoch": 3.594417663512468,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 15098
    },
    {
      "epoch": 3.5946557162411477,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 15099
    },
    {
      "epoch": 3.594893768969827,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 15100
    },
    {
      "epoch": 3.594893768969827,
      "eval_runtime": 0.5283,
      "eval_samples_per_second": 1.893,
      "eval_steps_per_second": 1.893,
      "step": 15100
    },
    {
      "epoch": 3.595131821698506,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 15101
    },
    {
      "epoch": 3.5953698744271856,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 15102
    },
    {
      "epoch": 3.595607927155865,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 15103
    },
    {
      "epoch": 3.5958459798845444,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 15104
    },
    {
      "epoch": 3.596084032613224,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 15105
    },
    {
      "epoch": 3.596322085341903,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 15106
    },
    {
      "epoch": 3.5965601380705827,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 15107
    },
    {
      "epoch": 3.596798190799262,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 15108
    },
    {
      "epoch": 3.5970362435279415,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 15109
    },
    {
      "epoch": 3.5972742962566207,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 15110
    },
    {
      "epoch": 3.5975123489853003,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 15111
    },
    {
      "epoch": 3.59775040171398,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.5793,
      "step": 15112
    },
    {
      "epoch": 3.597988454442659,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 15113
    },
    {
      "epoch": 3.598226507171338,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 15114
    },
    {
      "epoch": 3.598464559900018,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.5418,
      "step": 15115
    },
    {
      "epoch": 3.5987026126286974,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 15116
    },
    {
      "epoch": 3.5989406653573766,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4663,
      "step": 15117
    },
    {
      "epoch": 3.599178718086056,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 15118
    },
    {
      "epoch": 3.5994167708147353,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2561,
      "step": 15119
    },
    {
      "epoch": 3.599654823543415,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2604,
      "step": 15120
    },
    {
      "epoch": 3.5998928762720945,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 15121
    },
    {
      "epoch": 3.6001309290007737,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 15122
    },
    {
      "epoch": 3.600368981729453,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 15123
    },
    {
      "epoch": 3.6006070344581325,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 15124
    },
    {
      "epoch": 3.600845087186812,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 15125
    },
    {
      "epoch": 3.6010831399154912,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 15126
    },
    {
      "epoch": 3.601321192644171,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 15127
    },
    {
      "epoch": 3.60155924537285,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 15128
    },
    {
      "epoch": 3.6017972981015296,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 15129
    },
    {
      "epoch": 3.6020353508302088,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 15130
    },
    {
      "epoch": 3.6022734035588884,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 15131
    },
    {
      "epoch": 3.6025114562875675,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 15132
    },
    {
      "epoch": 3.602749509016247,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 15133
    },
    {
      "epoch": 3.6029875617449267,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 15134
    },
    {
      "epoch": 3.603225614473606,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 15135
    },
    {
      "epoch": 3.603463667202285,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 15136
    },
    {
      "epoch": 3.6037017199309647,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 15137
    },
    {
      "epoch": 3.6039397726596443,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 15138
    },
    {
      "epoch": 3.6041778253883234,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 15139
    },
    {
      "epoch": 3.604415878117003,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 15140
    },
    {
      "epoch": 3.604653930845682,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 15141
    },
    {
      "epoch": 3.604891983574362,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 15142
    },
    {
      "epoch": 3.605130036303041,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 15143
    },
    {
      "epoch": 3.6053680890317206,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 15144
    },
    {
      "epoch": 3.6056061417603997,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 15145
    },
    {
      "epoch": 3.6058441944890793,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 15146
    },
    {
      "epoch": 3.606082247217759,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 15147
    },
    {
      "epoch": 3.606320299946438,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4203,
      "step": 15148
    },
    {
      "epoch": 3.6065583526751173,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 15149
    },
    {
      "epoch": 3.606796405403797,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 15150
    },
    {
      "epoch": 3.6070344581324765,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3155,
      "step": 15151
    },
    {
      "epoch": 3.6072725108611556,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 15152
    },
    {
      "epoch": 3.6075105635898352,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2528,
      "step": 15153
    },
    {
      "epoch": 3.6077486163185144,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 15154
    },
    {
      "epoch": 3.607986669047194,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3161,
      "step": 15155
    },
    {
      "epoch": 3.6082247217758736,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8139,
      "step": 15156
    },
    {
      "epoch": 3.6084627745045528,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2809,
      "step": 15157
    },
    {
      "epoch": 3.608700827233232,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8308,
      "step": 15158
    },
    {
      "epoch": 3.6089388799619115,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 15159
    },
    {
      "epoch": 3.609176932690591,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 15160
    },
    {
      "epoch": 3.6094149854192703,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 15161
    },
    {
      "epoch": 3.60965303814795,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 15162
    },
    {
      "epoch": 3.609891090876629,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3416,
      "step": 15163
    },
    {
      "epoch": 3.6101291436053087,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 15164
    },
    {
      "epoch": 3.610367196333988,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 15165
    },
    {
      "epoch": 3.6106052490626674,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 15166
    },
    {
      "epoch": 3.6108433017913466,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 15167
    },
    {
      "epoch": 3.611081354520026,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3813,
      "step": 15168
    },
    {
      "epoch": 3.611319407248706,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 15169
    },
    {
      "epoch": 3.611557459977385,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2663,
      "step": 15170
    },
    {
      "epoch": 3.611795512706064,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 15171
    },
    {
      "epoch": 3.6120335654347437,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 15172
    },
    {
      "epoch": 3.6122716181634233,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 15173
    },
    {
      "epoch": 3.6125096708921025,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 15174
    },
    {
      "epoch": 3.612747723620782,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 15175
    },
    {
      "epoch": 3.6129857763494613,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 15176
    },
    {
      "epoch": 3.613223829078141,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 15177
    },
    {
      "epoch": 3.6134618818068205,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 15178
    },
    {
      "epoch": 3.6136999345354996,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.5268,
      "step": 15179
    },
    {
      "epoch": 3.613937987264179,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 15180
    },
    {
      "epoch": 3.6141760399928584,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.6499,
      "step": 15181
    },
    {
      "epoch": 3.614414092721538,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 15182
    },
    {
      "epoch": 3.614652145450217,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 15183
    },
    {
      "epoch": 3.614890198178897,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 15184
    },
    {
      "epoch": 3.615128250907576,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 15185
    },
    {
      "epoch": 3.6153663036362556,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 15186
    },
    {
      "epoch": 3.6156043563649347,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 15187
    },
    {
      "epoch": 3.6158424090936143,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 15188
    },
    {
      "epoch": 3.6160804618222935,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 15189
    },
    {
      "epoch": 3.616318514550973,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 15190
    },
    {
      "epoch": 3.6165565672796527,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 15191
    },
    {
      "epoch": 3.616794620008332,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 15192
    },
    {
      "epoch": 3.617032672737011,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.558,
      "step": 15193
    },
    {
      "epoch": 3.6172707254656906,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 15194
    },
    {
      "epoch": 3.6175087781943702,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 15195
    },
    {
      "epoch": 3.6177468309230494,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 15196
    },
    {
      "epoch": 3.617984883651729,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 15197
    },
    {
      "epoch": 3.618222936380408,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 15198
    },
    {
      "epoch": 3.6184609891090878,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 15199
    },
    {
      "epoch": 3.618699041837767,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 15200
    },
    {
      "epoch": 3.618699041837767,
      "eval_runtime": 0.528,
      "eval_samples_per_second": 1.894,
      "eval_steps_per_second": 1.894,
      "step": 15200
    },
    {
      "epoch": 3.6189370945664465,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 15201
    },
    {
      "epoch": 3.6191751472951257,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 15202
    },
    {
      "epoch": 3.6194132000238053,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3754,
      "step": 15203
    },
    {
      "epoch": 3.619651252752485,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 15204
    },
    {
      "epoch": 3.619889305481164,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 15205
    },
    {
      "epoch": 3.620127358209843,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 15206
    },
    {
      "epoch": 3.620365410938523,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 15207
    },
    {
      "epoch": 3.6206034636672024,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 15208
    },
    {
      "epoch": 3.6208415163958816,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 15209
    },
    {
      "epoch": 3.621079569124561,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3907,
      "step": 15210
    },
    {
      "epoch": 3.6213176218532404,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 15211
    },
    {
      "epoch": 3.62155567458192,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2796,
      "step": 15212
    },
    {
      "epoch": 3.6217937273105996,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 15213
    },
    {
      "epoch": 3.6220317800392787,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 15214
    },
    {
      "epoch": 3.622269832767958,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 15215
    },
    {
      "epoch": 3.6225078854966375,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 15216
    },
    {
      "epoch": 3.622745938225317,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 15217
    },
    {
      "epoch": 3.6229839909539963,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 15218
    },
    {
      "epoch": 3.623222043682676,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 15219
    },
    {
      "epoch": 3.623460096411355,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5194,
      "step": 15220
    },
    {
      "epoch": 3.6236981491400346,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 15221
    },
    {
      "epoch": 3.623936201868714,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 15222
    },
    {
      "epoch": 3.6241742545973934,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 15223
    },
    {
      "epoch": 3.6244123073260726,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 15224
    },
    {
      "epoch": 3.624650360054752,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 15225
    },
    {
      "epoch": 3.6248884127834318,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 15226
    },
    {
      "epoch": 3.625126465512111,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3464,
      "step": 15227
    },
    {
      "epoch": 3.62536451824079,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 15228
    },
    {
      "epoch": 3.6256025709694697,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 15229
    },
    {
      "epoch": 3.6258406236981493,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 15230
    },
    {
      "epoch": 3.6260786764268285,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 15231
    },
    {
      "epoch": 3.626316729155508,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 15232
    },
    {
      "epoch": 3.6265547818841872,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 15233
    },
    {
      "epoch": 3.626792834612867,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 15234
    },
    {
      "epoch": 3.627030887341546,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3841,
      "step": 15235
    },
    {
      "epoch": 3.6272689400702256,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 15236
    },
    {
      "epoch": 3.6275069927989048,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 15237
    },
    {
      "epoch": 3.6277450455275844,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 15238
    },
    {
      "epoch": 3.627983098256264,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 15239
    },
    {
      "epoch": 3.628221150984943,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 15240
    },
    {
      "epoch": 3.6284592037136223,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 15241
    },
    {
      "epoch": 3.628697256442302,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 15242
    },
    {
      "epoch": 3.6289353091709815,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 15243
    },
    {
      "epoch": 3.6291733618996607,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4719,
      "step": 15244
    },
    {
      "epoch": 3.6294114146283403,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 15245
    },
    {
      "epoch": 3.6296494673570194,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 15246
    },
    {
      "epoch": 3.629887520085699,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 15247
    },
    {
      "epoch": 3.6301255728143786,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 15248
    },
    {
      "epoch": 3.630363625543058,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 15249
    },
    {
      "epoch": 3.630601678271737,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 15250
    },
    {
      "epoch": 3.6308397310004166,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 15251
    },
    {
      "epoch": 3.631077783729096,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 15252
    },
    {
      "epoch": 3.6313158364577753,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 15253
    },
    {
      "epoch": 3.631553889186455,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5297,
      "step": 15254
    },
    {
      "epoch": 3.631791941915134,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 15255
    },
    {
      "epoch": 3.6320299946438137,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 15256
    },
    {
      "epoch": 3.632268047372493,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 15257
    },
    {
      "epoch": 3.6325061001011725,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 15258
    },
    {
      "epoch": 3.6327441528298516,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 15259
    },
    {
      "epoch": 3.6329822055585312,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 15260
    },
    {
      "epoch": 3.633220258287211,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 15261
    },
    {
      "epoch": 3.63345831101589,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 15262
    },
    {
      "epoch": 3.633696363744569,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 15263
    },
    {
      "epoch": 3.6339344164732488,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 15264
    },
    {
      "epoch": 3.6341724692019284,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 15265
    },
    {
      "epoch": 3.6344105219306075,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5567,
      "step": 15266
    },
    {
      "epoch": 3.634648574659287,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 15267
    },
    {
      "epoch": 3.6348866273879663,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 15268
    },
    {
      "epoch": 3.635124680116646,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 15269
    },
    {
      "epoch": 3.635362732845325,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 15270
    },
    {
      "epoch": 3.6356007855740047,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 15271
    },
    {
      "epoch": 3.635838838302684,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 15272
    },
    {
      "epoch": 3.6360768910313634,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 15273
    },
    {
      "epoch": 3.636314943760043,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 15274
    },
    {
      "epoch": 3.636552996488722,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.807,
      "step": 15275
    },
    {
      "epoch": 3.636791049217402,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 15276
    },
    {
      "epoch": 3.637029101946081,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 15277
    },
    {
      "epoch": 3.6372671546747606,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 15278
    },
    {
      "epoch": 3.6375052074034397,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2883,
      "step": 15279
    },
    {
      "epoch": 3.6377432601321193,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.378,
      "step": 15280
    },
    {
      "epoch": 3.6379813128607985,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 15281
    },
    {
      "epoch": 3.638219365589478,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 15282
    },
    {
      "epoch": 3.6384574183181577,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3044,
      "step": 15283
    },
    {
      "epoch": 3.638695471046837,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 15284
    },
    {
      "epoch": 3.638933523775516,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 15285
    },
    {
      "epoch": 3.6391715765041956,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 15286
    },
    {
      "epoch": 3.6394096292328753,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 15287
    },
    {
      "epoch": 3.6396476819615544,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 15288
    },
    {
      "epoch": 3.639885734690234,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3022,
      "step": 15289
    },
    {
      "epoch": 3.640123787418913,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 15290
    },
    {
      "epoch": 3.640361840147593,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 15291
    },
    {
      "epoch": 3.640599892876272,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 15292
    },
    {
      "epoch": 3.6408379456049516,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.41,
      "step": 15293
    },
    {
      "epoch": 3.6410759983336307,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 15294
    },
    {
      "epoch": 3.6413140510623103,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3575,
      "step": 15295
    },
    {
      "epoch": 3.64155210379099,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3113,
      "step": 15296
    },
    {
      "epoch": 3.641790156519669,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8815,
      "step": 15297
    },
    {
      "epoch": 3.6420282092483482,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.6007,
      "step": 15298
    },
    {
      "epoch": 3.642266261977028,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 15299
    },
    {
      "epoch": 3.6425043147057075,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 15300
    },
    {
      "epoch": 3.6425043147057075,
      "eval_runtime": 0.5187,
      "eval_samples_per_second": 1.928,
      "eval_steps_per_second": 1.928,
      "step": 15300
    },
    {
      "epoch": 3.6427423674343866,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 15301
    },
    {
      "epoch": 3.6429804201630662,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 15302
    },
    {
      "epoch": 3.6432184728917454,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 15303
    },
    {
      "epoch": 3.643456525620425,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 15304
    },
    {
      "epoch": 3.6436945783491046,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 15305
    },
    {
      "epoch": 3.6439326310777838,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 15306
    },
    {
      "epoch": 3.644170683806463,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 15307
    },
    {
      "epoch": 3.6444087365351425,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 15308
    },
    {
      "epoch": 3.644646789263822,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 15309
    },
    {
      "epoch": 3.6448848419925013,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 15310
    },
    {
      "epoch": 3.645122894721181,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3133,
      "step": 15311
    },
    {
      "epoch": 3.64536094744986,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 15312
    },
    {
      "epoch": 3.6455990001785397,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 15313
    },
    {
      "epoch": 3.645837052907219,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 15314
    },
    {
      "epoch": 3.6460751056358984,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 15315
    },
    {
      "epoch": 3.6463131583645776,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 15316
    },
    {
      "epoch": 3.646551211093257,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3286,
      "step": 15317
    },
    {
      "epoch": 3.646789263821937,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 15318
    },
    {
      "epoch": 3.647027316550616,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 15319
    },
    {
      "epoch": 3.647265369279295,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 15320
    },
    {
      "epoch": 3.6475034220079747,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 15321
    },
    {
      "epoch": 3.6477414747366543,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 15322
    },
    {
      "epoch": 3.6479795274653335,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 15323
    },
    {
      "epoch": 3.648217580194013,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2572,
      "step": 15324
    },
    {
      "epoch": 3.6484556329226923,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 15325
    },
    {
      "epoch": 3.648693685651372,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 15326
    },
    {
      "epoch": 3.648931738380051,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 15327
    },
    {
      "epoch": 3.6491697911087306,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 15328
    },
    {
      "epoch": 3.64940784383741,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 15329
    },
    {
      "epoch": 3.6496458965660894,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3704,
      "step": 15330
    },
    {
      "epoch": 3.649883949294769,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 15331
    },
    {
      "epoch": 3.650122002023448,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 15332
    },
    {
      "epoch": 3.6503600547521273,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 15333
    },
    {
      "epoch": 3.650598107480807,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 15334
    },
    {
      "epoch": 3.6508361602094865,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 15335
    },
    {
      "epoch": 3.6510742129381657,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 15336
    },
    {
      "epoch": 3.6513122656668453,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 15337
    },
    {
      "epoch": 3.6515503183955245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 15338
    },
    {
      "epoch": 3.651788371124204,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4419,
      "step": 15339
    },
    {
      "epoch": 3.6520264238528837,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 15340
    },
    {
      "epoch": 3.652264476581563,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 15341
    },
    {
      "epoch": 3.652502529310242,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 15342
    },
    {
      "epoch": 3.6527405820389216,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 15343
    },
    {
      "epoch": 3.652978634767601,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 15344
    },
    {
      "epoch": 3.6532166874962804,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 15345
    },
    {
      "epoch": 3.65345474022496,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 15346
    },
    {
      "epoch": 3.653692792953639,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 15347
    },
    {
      "epoch": 3.6539308456823187,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 15348
    },
    {
      "epoch": 3.654168898410998,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 15349
    },
    {
      "epoch": 3.6544069511396775,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 15350
    },
    {
      "epoch": 3.6546450038683567,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3647,
      "step": 15351
    },
    {
      "epoch": 3.6548830565970363,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 15352
    },
    {
      "epoch": 3.655121109325716,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 15353
    },
    {
      "epoch": 3.655359162054395,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 15354
    },
    {
      "epoch": 3.655597214783074,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 15355
    },
    {
      "epoch": 3.655835267511754,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 15356
    },
    {
      "epoch": 3.6560733202404334,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 15357
    },
    {
      "epoch": 3.6563113729691126,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5091,
      "step": 15358
    },
    {
      "epoch": 3.656549425697792,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 15359
    },
    {
      "epoch": 3.6567874784264713,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4188,
      "step": 15360
    },
    {
      "epoch": 3.657025531155151,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 15361
    },
    {
      "epoch": 3.65726358388383,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 15362
    },
    {
      "epoch": 3.6575016366125097,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 15363
    },
    {
      "epoch": 3.657739689341189,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2753,
      "step": 15364
    },
    {
      "epoch": 3.6579777420698685,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3564,
      "step": 15365
    },
    {
      "epoch": 3.658215794798548,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3448,
      "step": 15366
    },
    {
      "epoch": 3.6584538475272272,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3157,
      "step": 15367
    },
    {
      "epoch": 3.658691900255907,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 15368
    },
    {
      "epoch": 3.658929952984586,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 15369
    },
    {
      "epoch": 3.6591680057132656,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 15370
    },
    {
      "epoch": 3.6594060584419448,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2783,
      "step": 15371
    },
    {
      "epoch": 3.6596441111706244,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 15372
    },
    {
      "epoch": 3.6598821638993035,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 15373
    },
    {
      "epoch": 3.660120216627983,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 15374
    },
    {
      "epoch": 3.6603582693566628,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 15375
    },
    {
      "epoch": 3.660596322085342,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 15376
    },
    {
      "epoch": 3.660834374814021,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 15377
    },
    {
      "epoch": 3.6610724275427007,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 15378
    },
    {
      "epoch": 3.6613104802713803,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 15379
    },
    {
      "epoch": 3.6615485330000594,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4268,
      "step": 15380
    },
    {
      "epoch": 3.661786585728739,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 15381
    },
    {
      "epoch": 3.662024638457418,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 15382
    },
    {
      "epoch": 3.662262691186098,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 15383
    },
    {
      "epoch": 3.662500743914777,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 15384
    },
    {
      "epoch": 3.6627387966434566,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 15385
    },
    {
      "epoch": 3.6629768493721357,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 15386
    },
    {
      "epoch": 3.6632149021008154,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4152,
      "step": 15387
    },
    {
      "epoch": 3.663452954829495,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 15388
    },
    {
      "epoch": 3.663691007558174,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 15389
    },
    {
      "epoch": 3.6639290602868533,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 15390
    },
    {
      "epoch": 3.664167113015533,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 15391
    },
    {
      "epoch": 3.6644051657442125,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3946,
      "step": 15392
    },
    {
      "epoch": 3.6646432184728917,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 15393
    },
    {
      "epoch": 3.6648812712015713,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 15394
    },
    {
      "epoch": 3.6651193239302504,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 15395
    },
    {
      "epoch": 3.66535737665893,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 15396
    },
    {
      "epoch": 3.6655954293876096,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 15397
    },
    {
      "epoch": 3.665833482116289,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 15398
    },
    {
      "epoch": 3.666071534844968,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 15399
    },
    {
      "epoch": 3.6663095875736476,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 15400
    },
    {
      "epoch": 3.6663095875736476,
      "eval_runtime": 0.5182,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 15400
    },
    {
      "epoch": 3.666547640302327,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 15401
    },
    {
      "epoch": 3.6667856930310063,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4186,
      "step": 15402
    },
    {
      "epoch": 3.667023745759686,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 15403
    },
    {
      "epoch": 3.667261798488365,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 15404
    },
    {
      "epoch": 3.6674998512170447,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 15405
    },
    {
      "epoch": 3.667737903945724,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2961,
      "step": 15406
    },
    {
      "epoch": 3.6679759566744035,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 15407
    },
    {
      "epoch": 3.6682140094030826,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4804,
      "step": 15408
    },
    {
      "epoch": 3.6684520621317622,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 15409
    },
    {
      "epoch": 3.668690114860442,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 15410
    },
    {
      "epoch": 3.668928167589121,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 15411
    },
    {
      "epoch": 3.6691662203178,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 15412
    },
    {
      "epoch": 3.6694042730464798,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 15413
    },
    {
      "epoch": 3.6696423257751594,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5859,
      "step": 15414
    },
    {
      "epoch": 3.6698803785038385,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 15415
    },
    {
      "epoch": 3.670118431232518,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4551,
      "step": 15416
    },
    {
      "epoch": 3.6703564839611973,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 15417
    },
    {
      "epoch": 3.670594536689877,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 15418
    },
    {
      "epoch": 3.670832589418556,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 15419
    },
    {
      "epoch": 3.6710706421472357,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 15420
    },
    {
      "epoch": 3.671308694875915,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 15421
    },
    {
      "epoch": 3.6715467476045944,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 15422
    },
    {
      "epoch": 3.671784800333274,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 15423
    },
    {
      "epoch": 3.672022853061953,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 15424
    },
    {
      "epoch": 3.6722609057906324,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 15425
    },
    {
      "epoch": 3.672498958519312,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 15426
    },
    {
      "epoch": 3.6727370112479916,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 15427
    },
    {
      "epoch": 3.6729750639766707,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 15428
    },
    {
      "epoch": 3.6732131167053503,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3479,
      "step": 15429
    },
    {
      "epoch": 3.6734511694340295,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 15430
    },
    {
      "epoch": 3.673689222162709,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 15431
    },
    {
      "epoch": 3.6739272748913887,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 15432
    },
    {
      "epoch": 3.674165327620068,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2409,
      "step": 15433
    },
    {
      "epoch": 3.674403380348747,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 15434
    },
    {
      "epoch": 3.6746414330774266,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 15435
    },
    {
      "epoch": 3.6748794858061062,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 15436
    },
    {
      "epoch": 3.6751175385347854,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 15437
    },
    {
      "epoch": 3.675355591263465,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 15438
    },
    {
      "epoch": 3.675593643992144,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 15439
    },
    {
      "epoch": 3.6758316967208238,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 15440
    },
    {
      "epoch": 3.676069749449503,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 15441
    },
    {
      "epoch": 3.6763078021781825,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 15442
    },
    {
      "epoch": 3.6765458549068617,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 15443
    },
    {
      "epoch": 3.6767839076355413,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 15444
    },
    {
      "epoch": 3.677021960364221,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 15445
    },
    {
      "epoch": 3.6772600130929,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 15446
    },
    {
      "epoch": 3.6774980658215792,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3995,
      "step": 15447
    },
    {
      "epoch": 3.677736118550259,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 15448
    },
    {
      "epoch": 3.6779741712789384,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 15449
    },
    {
      "epoch": 3.6782122240076176,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 15450
    },
    {
      "epoch": 3.678450276736297,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 15451
    },
    {
      "epoch": 3.6786883294649764,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 15452
    },
    {
      "epoch": 3.678926382193656,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 15453
    },
    {
      "epoch": 3.679164434922335,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 15454
    },
    {
      "epoch": 3.6794024876510147,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 15455
    },
    {
      "epoch": 3.679640540379694,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.385,
      "step": 15456
    },
    {
      "epoch": 3.6798785931083735,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 15457
    },
    {
      "epoch": 3.680116645837053,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 15458
    },
    {
      "epoch": 3.6803546985657323,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 15459
    },
    {
      "epoch": 3.6805927512944114,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 15460
    },
    {
      "epoch": 3.680830804023091,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 15461
    },
    {
      "epoch": 3.6810688567517706,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 15462
    },
    {
      "epoch": 3.68130690948045,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 15463
    },
    {
      "epoch": 3.6815449622091294,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 15464
    },
    {
      "epoch": 3.6817830149378086,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 15465
    },
    {
      "epoch": 3.682021067666488,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 15466
    },
    {
      "epoch": 3.682259120395168,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 15467
    },
    {
      "epoch": 3.682497173123847,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 15468
    },
    {
      "epoch": 3.682735225852526,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 15469
    },
    {
      "epoch": 3.6829732785812057,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 15470
    },
    {
      "epoch": 3.6832113313098853,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 15471
    },
    {
      "epoch": 3.6834493840385645,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 15472
    },
    {
      "epoch": 3.683687436767244,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 15473
    },
    {
      "epoch": 3.6839254894959232,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 15474
    },
    {
      "epoch": 3.684163542224603,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 15475
    },
    {
      "epoch": 3.684401594953282,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 15476
    },
    {
      "epoch": 3.6846396476819616,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2695,
      "step": 15477
    },
    {
      "epoch": 3.6848777004106408,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 15478
    },
    {
      "epoch": 3.6851157531393204,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 15479
    },
    {
      "epoch": 3.685353805868,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 15480
    },
    {
      "epoch": 3.685591858596679,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 15481
    },
    {
      "epoch": 3.6858299113253583,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3674,
      "step": 15482
    },
    {
      "epoch": 3.686067964054038,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2555,
      "step": 15483
    },
    {
      "epoch": 3.6863060167827175,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 15484
    },
    {
      "epoch": 3.6865440695113967,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 15485
    },
    {
      "epoch": 3.6867821222400763,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2287,
      "step": 15486
    },
    {
      "epoch": 3.6870201749687554,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 15487
    },
    {
      "epoch": 3.687258227697435,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 15488
    },
    {
      "epoch": 3.6874962804261147,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 15489
    },
    {
      "epoch": 3.687734333154794,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 15490
    },
    {
      "epoch": 3.687972385883473,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 15491
    },
    {
      "epoch": 3.6882104386121526,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 15492
    },
    {
      "epoch": 3.688448491340832,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 15493
    },
    {
      "epoch": 3.6886865440695114,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 15494
    },
    {
      "epoch": 3.688924596798191,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8776,
      "step": 15495
    },
    {
      "epoch": 3.68916264952687,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 15496
    },
    {
      "epoch": 3.6894007022555497,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3987,
      "step": 15497
    },
    {
      "epoch": 3.689638754984229,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.396,
      "step": 15498
    },
    {
      "epoch": 3.6898768077129085,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 15499
    },
    {
      "epoch": 3.6901148604415877,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 15500
    },
    {
      "epoch": 3.6901148604415877,
      "eval_runtime": 0.5414,
      "eval_samples_per_second": 1.847,
      "eval_steps_per_second": 1.847,
      "step": 15500
    },
    {
      "epoch": 3.6903529131702673,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 15501
    },
    {
      "epoch": 3.690590965898947,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 15502
    },
    {
      "epoch": 3.690829018627626,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 15503
    },
    {
      "epoch": 3.691067071356305,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3861,
      "step": 15504
    },
    {
      "epoch": 3.691305124084985,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 15505
    },
    {
      "epoch": 3.6915431768136644,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 15506
    },
    {
      "epoch": 3.6917812295423436,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 15507
    },
    {
      "epoch": 3.692019282271023,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 15508
    },
    {
      "epoch": 3.6922573349997023,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 15509
    },
    {
      "epoch": 3.692495387728382,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3517,
      "step": 15510
    },
    {
      "epoch": 3.692733440457061,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 15511
    },
    {
      "epoch": 3.6929714931857407,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 15512
    },
    {
      "epoch": 3.69320954591442,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 15513
    },
    {
      "epoch": 3.6934475986430995,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 15514
    },
    {
      "epoch": 3.693685651371779,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2996,
      "step": 15515
    },
    {
      "epoch": 3.6939237041004582,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3025,
      "step": 15516
    },
    {
      "epoch": 3.6941617568291374,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 15517
    },
    {
      "epoch": 3.694399809557817,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 15518
    },
    {
      "epoch": 3.6946378622864966,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 15519
    },
    {
      "epoch": 3.6948759150151758,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 15520
    },
    {
      "epoch": 3.6951139677438554,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 15521
    },
    {
      "epoch": 3.6953520204725345,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 15522
    },
    {
      "epoch": 3.695590073201214,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 15523
    },
    {
      "epoch": 3.6958281259298937,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2942,
      "step": 15524
    },
    {
      "epoch": 3.696066178658573,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8293,
      "step": 15525
    },
    {
      "epoch": 3.696304231387252,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 15526
    },
    {
      "epoch": 3.6965422841159317,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 15527
    },
    {
      "epoch": 3.6967803368446113,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 15528
    },
    {
      "epoch": 3.6970183895732904,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3498,
      "step": 15529
    },
    {
      "epoch": 3.69725644230197,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 15530
    },
    {
      "epoch": 3.697494495030649,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 15531
    },
    {
      "epoch": 3.697732547759329,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 15532
    },
    {
      "epoch": 3.697970600488008,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 15533
    },
    {
      "epoch": 3.6982086532166876,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 15534
    },
    {
      "epoch": 3.6984467059453667,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 15535
    },
    {
      "epoch": 3.6986847586740463,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 15536
    },
    {
      "epoch": 3.698922811402726,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 15537
    },
    {
      "epoch": 3.699160864131405,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 15538
    },
    {
      "epoch": 3.6993989168600843,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 15539
    },
    {
      "epoch": 3.699636969588764,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 15540
    },
    {
      "epoch": 3.6998750223174435,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 15541
    },
    {
      "epoch": 3.7001130750461226,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 15542
    },
    {
      "epoch": 3.7003511277748022,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 15543
    },
    {
      "epoch": 3.7005891805034814,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 15544
    },
    {
      "epoch": 3.700827233232161,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 15545
    },
    {
      "epoch": 3.70106528596084,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 15546
    },
    {
      "epoch": 3.7013033386895198,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 15547
    },
    {
      "epoch": 3.701541391418199,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 15548
    },
    {
      "epoch": 3.7017794441468785,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 15549
    },
    {
      "epoch": 3.702017496875558,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 15550
    },
    {
      "epoch": 3.7022555496042373,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 15551
    },
    {
      "epoch": 3.7024936023329165,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 15552
    },
    {
      "epoch": 3.702731655061596,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 15553
    },
    {
      "epoch": 3.7029697077902757,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 15554
    },
    {
      "epoch": 3.703207760518955,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 15555
    },
    {
      "epoch": 3.7034458132476344,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 15556
    },
    {
      "epoch": 3.7036838659763136,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 15557
    },
    {
      "epoch": 3.703921918704993,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 15558
    },
    {
      "epoch": 3.704159971433673,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.284,
      "step": 15559
    },
    {
      "epoch": 3.704398024162352,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 15560
    },
    {
      "epoch": 3.704636076891031,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 15561
    },
    {
      "epoch": 3.7048741296197107,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 15562
    },
    {
      "epoch": 3.7051121823483903,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 15563
    },
    {
      "epoch": 3.7053502350770695,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 15564
    },
    {
      "epoch": 3.705588287805749,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 15565
    },
    {
      "epoch": 3.7058263405344283,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4621,
      "step": 15566
    },
    {
      "epoch": 3.706064393263108,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 15567
    },
    {
      "epoch": 3.706302445991787,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 15568
    },
    {
      "epoch": 3.7065404987204666,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 15569
    },
    {
      "epoch": 3.706778551449146,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 15570
    },
    {
      "epoch": 3.7070166041778254,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 15571
    },
    {
      "epoch": 3.707254656906505,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 15572
    },
    {
      "epoch": 3.707492709635184,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3822,
      "step": 15573
    },
    {
      "epoch": 3.7077307623638633,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 15574
    },
    {
      "epoch": 3.707968815092543,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 15575
    },
    {
      "epoch": 3.7082068678212226,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 15576
    },
    {
      "epoch": 3.7084449205499017,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 15577
    },
    {
      "epoch": 3.7086829732785813,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 15578
    },
    {
      "epoch": 3.7089210260072605,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4049,
      "step": 15579
    },
    {
      "epoch": 3.70915907873594,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3834,
      "step": 15580
    },
    {
      "epoch": 3.7093971314646197,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 15581
    },
    {
      "epoch": 3.709635184193299,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 15582
    },
    {
      "epoch": 3.709873236921978,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 15583
    },
    {
      "epoch": 3.7101112896506576,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 15584
    },
    {
      "epoch": 3.7103493423793372,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 15585
    },
    {
      "epoch": 3.7105873951080164,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 15586
    },
    {
      "epoch": 3.710825447836696,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2421,
      "step": 15587
    },
    {
      "epoch": 3.711063500565375,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 15588
    },
    {
      "epoch": 3.7113015532940548,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 15589
    },
    {
      "epoch": 3.711539606022734,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 15590
    },
    {
      "epoch": 3.7117776587514135,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 15591
    },
    {
      "epoch": 3.7120157114800927,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 15592
    },
    {
      "epoch": 3.7122537642087723,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 15593
    },
    {
      "epoch": 3.712491816937452,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 15594
    },
    {
      "epoch": 3.712729869666131,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 15595
    },
    {
      "epoch": 3.71296792239481,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 15596
    },
    {
      "epoch": 3.71320597512349,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 15597
    },
    {
      "epoch": 3.7134440278521694,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 15598
    },
    {
      "epoch": 3.7136820805808486,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 15599
    },
    {
      "epoch": 3.713920133309528,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 15600
    },
    {
      "epoch": 3.713920133309528,
      "eval_runtime": 0.5262,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 15600
    },
    {
      "epoch": 3.7141581860382074,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4724,
      "step": 15601
    },
    {
      "epoch": 3.714396238766887,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 15602
    },
    {
      "epoch": 3.714634291495566,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 15603
    },
    {
      "epoch": 3.7148723442242457,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 15604
    },
    {
      "epoch": 3.715110396952925,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 15605
    },
    {
      "epoch": 3.7153484496816045,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 15606
    },
    {
      "epoch": 3.715586502410284,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 15607
    },
    {
      "epoch": 3.7158245551389633,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 15608
    },
    {
      "epoch": 3.7160626078676424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 15609
    },
    {
      "epoch": 3.716300660596322,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 15610
    },
    {
      "epoch": 3.7165387133250016,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 15611
    },
    {
      "epoch": 3.716776766053681,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 15612
    },
    {
      "epoch": 3.7170148187823604,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 15613
    },
    {
      "epoch": 3.7172528715110396,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 15614
    },
    {
      "epoch": 3.717490924239719,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 15615
    },
    {
      "epoch": 3.7177289769683988,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 15616
    },
    {
      "epoch": 3.717967029697078,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 15617
    },
    {
      "epoch": 3.718205082425757,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 15618
    },
    {
      "epoch": 3.7184431351544367,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 15619
    },
    {
      "epoch": 3.7186811878831163,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 15620
    },
    {
      "epoch": 3.7189192406117955,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1948,
      "step": 15621
    },
    {
      "epoch": 3.719157293340475,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 15622
    },
    {
      "epoch": 3.7193953460691542,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 15623
    },
    {
      "epoch": 3.719633398797834,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 15624
    },
    {
      "epoch": 3.719871451526513,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 15625
    },
    {
      "epoch": 3.7201095042551926,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 15626
    },
    {
      "epoch": 3.7203475569838718,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 15627
    },
    {
      "epoch": 3.7205856097125514,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 15628
    },
    {
      "epoch": 3.720823662441231,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 15629
    },
    {
      "epoch": 3.72106171516991,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 15630
    },
    {
      "epoch": 3.7212997678985893,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 15631
    },
    {
      "epoch": 3.721537820627269,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5885,
      "step": 15632
    },
    {
      "epoch": 3.7217758733559485,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 15633
    },
    {
      "epoch": 3.7220139260846277,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.6002,
      "step": 15634
    },
    {
      "epoch": 3.7222519788133073,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 15635
    },
    {
      "epoch": 3.7224900315419864,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 15636
    },
    {
      "epoch": 3.722728084270666,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 15637
    },
    {
      "epoch": 3.722966136999345,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 15638
    },
    {
      "epoch": 3.723204189728025,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 15639
    },
    {
      "epoch": 3.723442242456704,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8404,
      "step": 15640
    },
    {
      "epoch": 3.7236802951853836,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 15641
    },
    {
      "epoch": 3.723918347914063,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 15642
    },
    {
      "epoch": 3.7241564006427423,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 15643
    },
    {
      "epoch": 3.7243944533714215,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 15644
    },
    {
      "epoch": 3.724632506100101,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 15645
    },
    {
      "epoch": 3.7248705588287807,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 15646
    },
    {
      "epoch": 3.72510861155746,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 15647
    },
    {
      "epoch": 3.7253466642861395,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 15648
    },
    {
      "epoch": 3.7255847170148186,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 15649
    },
    {
      "epoch": 3.7258227697434982,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 15650
    },
    {
      "epoch": 3.726060822472178,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 15651
    },
    {
      "epoch": 3.726298875200857,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3322,
      "step": 15652
    },
    {
      "epoch": 3.726536927929536,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 15653
    },
    {
      "epoch": 3.7267749806582158,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 15654
    },
    {
      "epoch": 3.7270130333868954,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 15655
    },
    {
      "epoch": 3.7272510861155745,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3587,
      "step": 15656
    },
    {
      "epoch": 3.727489138844254,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 15657
    },
    {
      "epoch": 3.7277271915729333,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 15658
    },
    {
      "epoch": 3.727965244301613,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 15659
    },
    {
      "epoch": 3.728203297030292,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3296,
      "step": 15660
    },
    {
      "epoch": 3.7284413497589717,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 15661
    },
    {
      "epoch": 3.728679402487651,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 15662
    },
    {
      "epoch": 3.7289174552163304,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 15663
    },
    {
      "epoch": 3.72915550794501,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 15664
    },
    {
      "epoch": 3.729393560673689,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 15665
    },
    {
      "epoch": 3.7296316134023684,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 15666
    },
    {
      "epoch": 3.729869666131048,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 15667
    },
    {
      "epoch": 3.7301077188597276,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 15668
    },
    {
      "epoch": 3.7303457715884067,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 15669
    },
    {
      "epoch": 3.7305838243170863,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 15670
    },
    {
      "epoch": 3.7308218770457655,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4913,
      "step": 15671
    },
    {
      "epoch": 3.731059929774445,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 15672
    },
    {
      "epoch": 3.7312979825031243,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 15673
    },
    {
      "epoch": 3.731536035231804,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 15674
    },
    {
      "epoch": 3.731774087960483,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 15675
    },
    {
      "epoch": 3.7320121406891626,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.545,
      "step": 15676
    },
    {
      "epoch": 3.7322501934178423,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 15677
    },
    {
      "epoch": 3.7324882461465214,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 15678
    },
    {
      "epoch": 3.732726298875201,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 15679
    },
    {
      "epoch": 3.73296435160388,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 15680
    },
    {
      "epoch": 3.73320240433256,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 15681
    },
    {
      "epoch": 3.733440457061239,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 15682
    },
    {
      "epoch": 3.7336785097899186,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 15683
    },
    {
      "epoch": 3.7339165625185977,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4087,
      "step": 15684
    },
    {
      "epoch": 3.7341546152472773,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 15685
    },
    {
      "epoch": 3.734392667975957,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 15686
    },
    {
      "epoch": 3.734630720704636,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 15687
    },
    {
      "epoch": 3.7348687734333152,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 15688
    },
    {
      "epoch": 3.735106826161995,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 15689
    },
    {
      "epoch": 3.7353448788906745,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.492,
      "step": 15690
    },
    {
      "epoch": 3.7355829316193536,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2969,
      "step": 15691
    },
    {
      "epoch": 3.7358209843480332,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4575,
      "step": 15692
    },
    {
      "epoch": 3.7360590370767124,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 15693
    },
    {
      "epoch": 3.736297089805392,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.4091,
      "step": 15694
    },
    {
      "epoch": 3.736535142534071,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 15695
    },
    {
      "epoch": 3.7367731952627508,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 15696
    },
    {
      "epoch": 3.73701124799143,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 15697
    },
    {
      "epoch": 3.7372493007201095,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 15698
    },
    {
      "epoch": 3.737487353448789,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 15699
    },
    {
      "epoch": 3.7377254061774683,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 15700
    },
    {
      "epoch": 3.7377254061774683,
      "eval_runtime": 0.5207,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 15700
    },
    {
      "epoch": 3.7379634589061475,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 15701
    },
    {
      "epoch": 3.738201511634827,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 15702
    },
    {
      "epoch": 3.7384395643635067,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 15703
    },
    {
      "epoch": 3.738677617092186,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 15704
    },
    {
      "epoch": 3.7389156698208654,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 15705
    },
    {
      "epoch": 3.7391537225495446,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3245,
      "step": 15706
    },
    {
      "epoch": 3.739391775278224,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 15707
    },
    {
      "epoch": 3.739629828006904,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 15708
    },
    {
      "epoch": 3.739867880735583,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 15709
    },
    {
      "epoch": 3.740105933464262,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 15710
    },
    {
      "epoch": 3.7403439861929417,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 15711
    },
    {
      "epoch": 3.7405820389216213,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 15712
    },
    {
      "epoch": 3.7408200916503005,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 15713
    },
    {
      "epoch": 3.74105814437898,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 15714
    },
    {
      "epoch": 3.7412961971076593,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 15715
    },
    {
      "epoch": 3.741534249836339,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 15716
    },
    {
      "epoch": 3.741772302565018,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 15717
    },
    {
      "epoch": 3.7420103552936976,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 15718
    },
    {
      "epoch": 3.742248408022377,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 15719
    },
    {
      "epoch": 3.7424864607510564,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 15720
    },
    {
      "epoch": 3.742724513479736,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 15721
    },
    {
      "epoch": 3.742962566208415,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 15722
    },
    {
      "epoch": 3.7432006189370943,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 15723
    },
    {
      "epoch": 3.743438671665774,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 15724
    },
    {
      "epoch": 3.7436767243944535,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 15725
    },
    {
      "epoch": 3.7439147771231327,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 15726
    },
    {
      "epoch": 3.7441528298518123,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 15727
    },
    {
      "epoch": 3.7443908825804915,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3416,
      "step": 15728
    },
    {
      "epoch": 3.744628935309171,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4684,
      "step": 15729
    },
    {
      "epoch": 3.7448669880378502,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 15730
    },
    {
      "epoch": 3.74510504076653,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8406,
      "step": 15731
    },
    {
      "epoch": 3.745343093495209,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 15732
    },
    {
      "epoch": 3.7455811462238886,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 15733
    },
    {
      "epoch": 3.745819198952568,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 15734
    },
    {
      "epoch": 3.7460572516812474,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 15735
    },
    {
      "epoch": 3.7462953044099265,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 15736
    },
    {
      "epoch": 3.746533357138606,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 15737
    },
    {
      "epoch": 3.7467714098672857,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 15738
    },
    {
      "epoch": 3.747009462595965,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4567,
      "step": 15739
    },
    {
      "epoch": 3.7472475153246445,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 15740
    },
    {
      "epoch": 3.7474855680533237,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 15741
    },
    {
      "epoch": 3.7477236207820033,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 15742
    },
    {
      "epoch": 3.747961673510683,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 15743
    },
    {
      "epoch": 3.748199726239362,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 15744
    },
    {
      "epoch": 3.748437778968041,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 15745
    },
    {
      "epoch": 3.748675831696721,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 15746
    },
    {
      "epoch": 3.7489138844254004,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3884,
      "step": 15747
    },
    {
      "epoch": 3.7491519371540796,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 15748
    },
    {
      "epoch": 3.749389989882759,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 15749
    },
    {
      "epoch": 3.7496280426114383,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 15750
    },
    {
      "epoch": 3.749866095340118,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 15751
    },
    {
      "epoch": 3.750104148068797,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 15752
    },
    {
      "epoch": 3.7503422007974767,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 15753
    },
    {
      "epoch": 3.750580253526156,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 15754
    },
    {
      "epoch": 3.7508183062548355,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 15755
    },
    {
      "epoch": 3.751056358983515,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.774,
      "step": 15756
    },
    {
      "epoch": 3.7512944117121942,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 15757
    },
    {
      "epoch": 3.7515324644408734,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 15758
    },
    {
      "epoch": 3.751770517169553,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 15759
    },
    {
      "epoch": 3.7520085698982326,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 15760
    },
    {
      "epoch": 3.7522466226269118,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 15761
    },
    {
      "epoch": 3.7524846753555914,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 15762
    },
    {
      "epoch": 3.7527227280842705,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 15763
    },
    {
      "epoch": 3.75296078081295,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 15764
    },
    {
      "epoch": 3.7531988335416293,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 15765
    },
    {
      "epoch": 3.753436886270309,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 15766
    },
    {
      "epoch": 3.753674938998988,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 15767
    },
    {
      "epoch": 3.7539129917276677,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2877,
      "step": 15768
    },
    {
      "epoch": 3.7541510444563473,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3058,
      "step": 15769
    },
    {
      "epoch": 3.7543890971850264,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 15770
    },
    {
      "epoch": 3.754627149913706,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3017,
      "step": 15771
    },
    {
      "epoch": 3.754865202642385,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3637,
      "step": 15772
    },
    {
      "epoch": 3.755103255371065,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 15773
    },
    {
      "epoch": 3.755341308099744,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 15774
    },
    {
      "epoch": 3.7555793608284236,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 15775
    },
    {
      "epoch": 3.7558174135571027,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 15776
    },
    {
      "epoch": 3.7560554662857824,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 15777
    },
    {
      "epoch": 3.756293519014462,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.6332,
      "step": 15778
    },
    {
      "epoch": 3.756531571743141,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 15779
    },
    {
      "epoch": 3.7567696244718203,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 15780
    },
    {
      "epoch": 3.7570076772005,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 15781
    },
    {
      "epoch": 3.7572457299291795,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.31,
      "step": 15782
    },
    {
      "epoch": 3.7574837826578587,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 15783
    },
    {
      "epoch": 3.7577218353865383,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 15784
    },
    {
      "epoch": 3.7579598881152174,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3459,
      "step": 15785
    },
    {
      "epoch": 3.758197940843897,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7264,
      "step": 15786
    },
    {
      "epoch": 3.758435993572576,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 15787
    },
    {
      "epoch": 3.758674046301256,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 15788
    },
    {
      "epoch": 3.758912099029935,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 15789
    },
    {
      "epoch": 3.7591501517586146,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 15790
    },
    {
      "epoch": 3.759388204487294,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4009,
      "step": 15791
    },
    {
      "epoch": 3.7596262572159733,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 15792
    },
    {
      "epoch": 3.7598643099446525,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 15793
    },
    {
      "epoch": 3.760102362673332,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 15794
    },
    {
      "epoch": 3.7603404154020117,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3136,
      "step": 15795
    },
    {
      "epoch": 3.760578468130691,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 15796
    },
    {
      "epoch": 3.7608165208593705,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2004,
      "step": 15797
    },
    {
      "epoch": 3.7610545735880496,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 15798
    },
    {
      "epoch": 3.7612926263167292,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 15799
    },
    {
      "epoch": 3.761530679045409,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2811,
      "step": 15800
    },
    {
      "epoch": 3.761530679045409,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 15800
    },
    {
      "epoch": 3.761768731774088,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 15801
    },
    {
      "epoch": 3.762006784502767,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 15802
    },
    {
      "epoch": 3.7622448372314468,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 15803
    },
    {
      "epoch": 3.7624828899601264,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 15804
    },
    {
      "epoch": 3.7627209426888055,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 15805
    },
    {
      "epoch": 3.762958995417485,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 15806
    },
    {
      "epoch": 3.7631970481461643,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 15807
    },
    {
      "epoch": 3.763435100874844,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 15808
    },
    {
      "epoch": 3.763673153603523,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 15809
    },
    {
      "epoch": 3.7639112063322027,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 15810
    },
    {
      "epoch": 3.764149259060882,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 15811
    },
    {
      "epoch": 3.7643873117895614,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 15812
    },
    {
      "epoch": 3.764625364518241,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 15813
    },
    {
      "epoch": 3.76486341724692,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 15814
    },
    {
      "epoch": 3.7651014699755994,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 15815
    },
    {
      "epoch": 3.765339522704279,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 15816
    },
    {
      "epoch": 3.7655775754329586,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 15817
    },
    {
      "epoch": 3.7658156281616377,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 15818
    },
    {
      "epoch": 3.7660536808903173,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 15819
    },
    {
      "epoch": 3.7662917336189965,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 15820
    },
    {
      "epoch": 3.766529786347676,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 15821
    },
    {
      "epoch": 3.7667678390763553,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3971,
      "step": 15822
    },
    {
      "epoch": 3.767005891805035,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.388,
      "step": 15823
    },
    {
      "epoch": 3.767243944533714,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 15824
    },
    {
      "epoch": 3.7674819972623936,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 15825
    },
    {
      "epoch": 3.7677200499910732,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 15826
    },
    {
      "epoch": 3.7679581027197524,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 15827
    },
    {
      "epoch": 3.7681961554484316,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 15828
    },
    {
      "epoch": 3.768434208177111,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 15829
    },
    {
      "epoch": 3.7686722609057908,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 15830
    },
    {
      "epoch": 3.76891031363447,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 15831
    },
    {
      "epoch": 3.7691483663631495,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 15832
    },
    {
      "epoch": 3.7693864190918287,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 15833
    },
    {
      "epoch": 3.7696244718205083,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 15834
    },
    {
      "epoch": 3.769862524549188,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3333,
      "step": 15835
    },
    {
      "epoch": 3.770100577277867,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 15836
    },
    {
      "epoch": 3.7703386300065462,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 15837
    },
    {
      "epoch": 3.770576682735226,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 15838
    },
    {
      "epoch": 3.7708147354639054,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 15839
    },
    {
      "epoch": 3.7710527881925846,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 15840
    },
    {
      "epoch": 3.771290840921264,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 15841
    },
    {
      "epoch": 3.7715288936499434,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 15842
    },
    {
      "epoch": 3.771766946378623,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 15843
    },
    {
      "epoch": 3.772004999107302,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7711,
      "step": 15844
    },
    {
      "epoch": 3.7722430518359817,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 15845
    },
    {
      "epoch": 3.772481104564661,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 15846
    },
    {
      "epoch": 3.7727191572933405,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 15847
    },
    {
      "epoch": 3.77295721002202,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 15848
    },
    {
      "epoch": 3.7731952627506993,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.504,
      "step": 15849
    },
    {
      "epoch": 3.7734333154793784,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 15850
    },
    {
      "epoch": 3.773671368208058,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 15851
    },
    {
      "epoch": 3.7739094209367376,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 15852
    },
    {
      "epoch": 3.774147473665417,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 15853
    },
    {
      "epoch": 3.7743855263940964,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 15854
    },
    {
      "epoch": 3.7746235791227756,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 15855
    },
    {
      "epoch": 3.774861631851455,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 15856
    },
    {
      "epoch": 3.7750996845801343,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 15857
    },
    {
      "epoch": 3.775337737308814,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 15858
    },
    {
      "epoch": 3.775575790037493,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 15859
    },
    {
      "epoch": 3.7758138427661727,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4639,
      "step": 15860
    },
    {
      "epoch": 3.7760518954948523,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 15861
    },
    {
      "epoch": 3.7762899482235315,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 15862
    },
    {
      "epoch": 3.7765280009522106,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 15863
    },
    {
      "epoch": 3.7767660536808902,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 15864
    },
    {
      "epoch": 3.77700410640957,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 15865
    },
    {
      "epoch": 3.777242159138249,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5312,
      "step": 15866
    },
    {
      "epoch": 3.7774802118669286,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3093,
      "step": 15867
    },
    {
      "epoch": 3.7777182645956078,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 15868
    },
    {
      "epoch": 3.7779563173242874,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 15869
    },
    {
      "epoch": 3.778194370052967,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3748,
      "step": 15870
    },
    {
      "epoch": 3.778432422781646,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 15871
    },
    {
      "epoch": 3.7786704755103253,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 15872
    },
    {
      "epoch": 3.778908528239005,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 15873
    },
    {
      "epoch": 3.7791465809676845,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4695,
      "step": 15874
    },
    {
      "epoch": 3.7793846336963637,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 15875
    },
    {
      "epoch": 3.7796226864250433,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 15876
    },
    {
      "epoch": 3.7798607391537224,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 15877
    },
    {
      "epoch": 3.780098791882402,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3256,
      "step": 15878
    },
    {
      "epoch": 3.780336844611081,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 15879
    },
    {
      "epoch": 3.780574897339761,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 15880
    },
    {
      "epoch": 3.78081295006844,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 15881
    },
    {
      "epoch": 3.7810510027971196,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 15882
    },
    {
      "epoch": 3.781289055525799,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 15883
    },
    {
      "epoch": 3.7815271082544784,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 15884
    },
    {
      "epoch": 3.7817651609831575,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 15885
    },
    {
      "epoch": 3.782003213711837,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 15886
    },
    {
      "epoch": 3.7822412664405167,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 15887
    },
    {
      "epoch": 3.782479319169196,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 15888
    },
    {
      "epoch": 3.7827173718978755,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 15889
    },
    {
      "epoch": 3.7829554246265547,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 15890
    },
    {
      "epoch": 3.7831934773552343,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 15891
    },
    {
      "epoch": 3.783431530083914,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 15892
    },
    {
      "epoch": 3.783669582812593,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 15893
    },
    {
      "epoch": 3.783907635541272,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 15894
    },
    {
      "epoch": 3.784145688269952,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2789,
      "step": 15895
    },
    {
      "epoch": 3.7843837409986314,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 15896
    },
    {
      "epoch": 3.7846217937273106,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 15897
    },
    {
      "epoch": 3.78485984645599,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 15898
    },
    {
      "epoch": 3.7850978991846693,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3358,
      "step": 15899
    },
    {
      "epoch": 3.785335951913349,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 15900
    },
    {
      "epoch": 3.785335951913349,
      "eval_runtime": 0.5254,
      "eval_samples_per_second": 1.903,
      "eval_steps_per_second": 1.903,
      "step": 15900
    },
    {
      "epoch": 3.785574004642028,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 15901
    },
    {
      "epoch": 3.7858120573707077,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 15902
    },
    {
      "epoch": 3.786050110099387,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 15903
    },
    {
      "epoch": 3.7862881628280665,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4446,
      "step": 15904
    },
    {
      "epoch": 3.786526215556746,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 15905
    },
    {
      "epoch": 3.7867642682854252,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 15906
    },
    {
      "epoch": 3.7870023210141044,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 15907
    },
    {
      "epoch": 3.787240373742784,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 15908
    },
    {
      "epoch": 3.7874784264714636,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 15909
    },
    {
      "epoch": 3.7877164792001428,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 15910
    },
    {
      "epoch": 3.7879545319288224,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 15911
    },
    {
      "epoch": 3.7881925846575015,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 15912
    },
    {
      "epoch": 3.788430637386181,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 15913
    },
    {
      "epoch": 3.7886686901148603,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 15914
    },
    {
      "epoch": 3.78890674284354,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 15915
    },
    {
      "epoch": 3.789144795572219,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 15916
    },
    {
      "epoch": 3.7893828483008987,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 15917
    },
    {
      "epoch": 3.7896209010295783,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3117,
      "step": 15918
    },
    {
      "epoch": 3.7898589537582574,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 15919
    },
    {
      "epoch": 3.7900970064869366,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 15920
    },
    {
      "epoch": 3.790335059215616,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 15921
    },
    {
      "epoch": 3.790573111944296,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 15922
    },
    {
      "epoch": 3.790811164672975,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 15923
    },
    {
      "epoch": 3.7910492174016546,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 15924
    },
    {
      "epoch": 3.7912872701303337,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 15925
    },
    {
      "epoch": 3.7915253228590133,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1914,
      "step": 15926
    },
    {
      "epoch": 3.791763375587693,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 15927
    },
    {
      "epoch": 3.792001428316372,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 15928
    },
    {
      "epoch": 3.7922394810450513,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 15929
    },
    {
      "epoch": 3.792477533773731,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 15930
    },
    {
      "epoch": 3.7927155865024105,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 15931
    },
    {
      "epoch": 3.7929536392310896,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 15932
    },
    {
      "epoch": 3.7931916919597692,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4063,
      "step": 15933
    },
    {
      "epoch": 3.7934297446884484,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4108,
      "step": 15934
    },
    {
      "epoch": 3.793667797417128,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.5232,
      "step": 15935
    },
    {
      "epoch": 3.793905850145807,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 15936
    },
    {
      "epoch": 3.7941439028744868,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 15937
    },
    {
      "epoch": 3.794381955603166,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 15938
    },
    {
      "epoch": 3.7946200083318455,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 15939
    },
    {
      "epoch": 3.794858061060525,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 15940
    },
    {
      "epoch": 3.7950961137892043,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 15941
    },
    {
      "epoch": 3.7953341665178835,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 15942
    },
    {
      "epoch": 3.795572219246563,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 15943
    },
    {
      "epoch": 3.7958102719752427,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 15944
    },
    {
      "epoch": 3.796048324703922,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 15945
    },
    {
      "epoch": 3.7962863774326014,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 15946
    },
    {
      "epoch": 3.7965244301612806,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 15947
    },
    {
      "epoch": 3.79676248288996,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 15948
    },
    {
      "epoch": 3.7970005356186394,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 15949
    },
    {
      "epoch": 3.797238588347319,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 15950
    },
    {
      "epoch": 3.797476641075998,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 15951
    },
    {
      "epoch": 3.7977146938046777,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 15952
    },
    {
      "epoch": 3.7979527465333573,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 15953
    },
    {
      "epoch": 3.7981907992620365,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 15954
    },
    {
      "epoch": 3.7984288519907157,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 15955
    },
    {
      "epoch": 3.7986669047193953,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 15956
    },
    {
      "epoch": 3.798904957448075,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 15957
    },
    {
      "epoch": 3.799143010176754,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 15958
    },
    {
      "epoch": 3.7993810629054336,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 15959
    },
    {
      "epoch": 3.799619115634113,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 15960
    },
    {
      "epoch": 3.7998571683627924,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 15961
    },
    {
      "epoch": 3.800095221091472,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 15962
    },
    {
      "epoch": 3.800333273820151,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3207,
      "step": 15963
    },
    {
      "epoch": 3.8005713265488303,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 15964
    },
    {
      "epoch": 3.80080937927751,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3604,
      "step": 15965
    },
    {
      "epoch": 3.8010474320061896,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 15966
    },
    {
      "epoch": 3.8012854847348687,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 15967
    },
    {
      "epoch": 3.8015235374635483,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 15968
    },
    {
      "epoch": 3.8017615901922275,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 15969
    },
    {
      "epoch": 3.801999642920907,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 15970
    },
    {
      "epoch": 3.8022376956495862,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 15971
    },
    {
      "epoch": 3.802475748378266,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8434,
      "step": 15972
    },
    {
      "epoch": 3.802713801106945,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4932,
      "step": 15973
    },
    {
      "epoch": 3.8029518538356246,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 15974
    },
    {
      "epoch": 3.8031899065643042,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 15975
    },
    {
      "epoch": 3.8034279592929834,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 15976
    },
    {
      "epoch": 3.8036660120216625,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 15977
    },
    {
      "epoch": 3.803904064750342,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 15978
    },
    {
      "epoch": 3.8041421174790218,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 15979
    },
    {
      "epoch": 3.804380170207701,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 15980
    },
    {
      "epoch": 3.8046182229363805,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 15981
    },
    {
      "epoch": 3.8048562756650597,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 15982
    },
    {
      "epoch": 3.8050943283937393,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 15983
    },
    {
      "epoch": 3.8053323811224185,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 15984
    },
    {
      "epoch": 3.805570433851098,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 15985
    },
    {
      "epoch": 3.805808486579777,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 15986
    },
    {
      "epoch": 3.806046539308457,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 15987
    },
    {
      "epoch": 3.8062845920371364,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 15988
    },
    {
      "epoch": 3.8065226447658156,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 15989
    },
    {
      "epoch": 3.806760697494495,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 15990
    },
    {
      "epoch": 3.8069987502231744,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 15991
    },
    {
      "epoch": 3.807236802951854,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 15992
    },
    {
      "epoch": 3.807474855680533,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 15993
    },
    {
      "epoch": 3.8077129084092127,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 15994
    },
    {
      "epoch": 3.807950961137892,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 15995
    },
    {
      "epoch": 3.8081890138665715,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.665,
      "step": 15996
    },
    {
      "epoch": 3.808427066595251,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 15997
    },
    {
      "epoch": 3.8086651193239303,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 15998
    },
    {
      "epoch": 3.8089031720526094,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 15999
    },
    {
      "epoch": 3.809141224781289,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 16000
    },
    {
      "epoch": 3.809141224781289,
      "eval_runtime": 0.5651,
      "eval_samples_per_second": 1.77,
      "eval_steps_per_second": 1.77,
      "step": 16000
    },
    {
      "epoch": 3.8093792775099686,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3576,
      "step": 16001
    },
    {
      "epoch": 3.809617330238648,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2439,
      "step": 16002
    },
    {
      "epoch": 3.8098553829673274,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3902,
      "step": 16003
    },
    {
      "epoch": 3.8100934356960066,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 16004
    },
    {
      "epoch": 3.810331488424686,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 16005
    },
    {
      "epoch": 3.8105695411533653,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 16006
    },
    {
      "epoch": 3.810807593882045,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.436,
      "step": 16007
    },
    {
      "epoch": 3.811045646610724,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 16008
    },
    {
      "epoch": 3.8112836993394037,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 16009
    },
    {
      "epoch": 3.8115217520680833,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3307,
      "step": 16010
    },
    {
      "epoch": 3.8117598047967625,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 16011
    },
    {
      "epoch": 3.8119978575254416,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 16012
    },
    {
      "epoch": 3.8122359102541212,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 16013
    },
    {
      "epoch": 3.812473962982801,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 16014
    },
    {
      "epoch": 3.81271201571148,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 16015
    },
    {
      "epoch": 3.8129500684401596,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 16016
    },
    {
      "epoch": 3.8131881211688388,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 16017
    },
    {
      "epoch": 3.8134261738975184,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3768,
      "step": 16018
    },
    {
      "epoch": 3.813664226626198,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 16019
    },
    {
      "epoch": 3.813902279354877,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4211,
      "step": 16020
    },
    {
      "epoch": 3.8141403320835563,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 16021
    },
    {
      "epoch": 3.814378384812236,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 16022
    },
    {
      "epoch": 3.8146164375409155,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 16023
    },
    {
      "epoch": 3.8148544902695947,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 16024
    },
    {
      "epoch": 3.8150925429982743,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 16025
    },
    {
      "epoch": 3.8153305957269534,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4563,
      "step": 16026
    },
    {
      "epoch": 3.815568648455633,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 16027
    },
    {
      "epoch": 3.815806701184312,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 16028
    },
    {
      "epoch": 3.816044753912992,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 16029
    },
    {
      "epoch": 3.816282806641671,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.933,
      "step": 16030
    },
    {
      "epoch": 3.8165208593703506,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 16031
    },
    {
      "epoch": 3.81675891209903,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 16032
    },
    {
      "epoch": 3.8169969648277093,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 16033
    },
    {
      "epoch": 3.8172350175563885,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 16034
    },
    {
      "epoch": 3.817473070285068,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4791,
      "step": 16035
    },
    {
      "epoch": 3.8177111230137477,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 16036
    },
    {
      "epoch": 3.817949175742427,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 16037
    },
    {
      "epoch": 3.8181872284711065,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 16038
    },
    {
      "epoch": 3.8184252811997856,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 16039
    },
    {
      "epoch": 3.8186633339284652,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 16040
    },
    {
      "epoch": 3.8189013866571444,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 16041
    },
    {
      "epoch": 3.819139439385824,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 16042
    },
    {
      "epoch": 3.819377492114503,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 16043
    },
    {
      "epoch": 3.8196155448431828,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 16044
    },
    {
      "epoch": 3.8198535975718624,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 16045
    },
    {
      "epoch": 3.8200916503005415,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 16046
    },
    {
      "epoch": 3.8203297030292207,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 16047
    },
    {
      "epoch": 3.8205677557579003,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 16048
    },
    {
      "epoch": 3.82080580848658,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 16049
    },
    {
      "epoch": 3.821043861215259,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3759,
      "step": 16050
    },
    {
      "epoch": 3.8212819139439387,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 16051
    },
    {
      "epoch": 3.821519966672618,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 16052
    },
    {
      "epoch": 3.8217580194012974,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 16053
    },
    {
      "epoch": 3.821996072129977,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 16054
    },
    {
      "epoch": 3.822234124858656,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 16055
    },
    {
      "epoch": 3.8224721775873354,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 16056
    },
    {
      "epoch": 3.822710230316015,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2693,
      "step": 16057
    },
    {
      "epoch": 3.8229482830446946,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 16058
    },
    {
      "epoch": 3.8231863357733737,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 16059
    },
    {
      "epoch": 3.8234243885020534,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 16060
    },
    {
      "epoch": 3.8236624412307325,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 16061
    },
    {
      "epoch": 3.823900493959412,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 16062
    },
    {
      "epoch": 3.8241385466880913,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 16063
    },
    {
      "epoch": 3.824376599416771,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3704,
      "step": 16064
    },
    {
      "epoch": 3.82461465214545,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3893,
      "step": 16065
    },
    {
      "epoch": 3.8248527048741296,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 16066
    },
    {
      "epoch": 3.8250907576028093,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 16067
    },
    {
      "epoch": 3.8253288103314884,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 16068
    },
    {
      "epoch": 3.8255668630601676,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 16069
    },
    {
      "epoch": 3.825804915788847,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 16070
    },
    {
      "epoch": 3.826042968517527,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 16071
    },
    {
      "epoch": 3.826281021246206,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 16072
    },
    {
      "epoch": 3.8265190739748856,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 16073
    },
    {
      "epoch": 3.8267571267035647,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 16074
    },
    {
      "epoch": 3.8269951794322443,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3247,
      "step": 16075
    },
    {
      "epoch": 3.8272332321609235,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 16076
    },
    {
      "epoch": 3.827471284889603,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 16077
    },
    {
      "epoch": 3.8277093376182822,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 16078
    },
    {
      "epoch": 3.827947390346962,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 16079
    },
    {
      "epoch": 3.8281854430756415,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 16080
    },
    {
      "epoch": 3.8284234958043206,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 16081
    },
    {
      "epoch": 3.8286615485330002,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 16082
    },
    {
      "epoch": 3.8288996012616794,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 16083
    },
    {
      "epoch": 3.829137653990359,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 16084
    },
    {
      "epoch": 3.829375706719038,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 16085
    },
    {
      "epoch": 3.8296137594477178,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 16086
    },
    {
      "epoch": 3.829851812176397,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 16087
    },
    {
      "epoch": 3.8300898649050765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 16088
    },
    {
      "epoch": 3.830327917633756,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 16089
    },
    {
      "epoch": 3.8305659703624353,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 16090
    },
    {
      "epoch": 3.8308040230911145,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 16091
    },
    {
      "epoch": 3.831042075819794,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 16092
    },
    {
      "epoch": 3.8312801285484737,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 16093
    },
    {
      "epoch": 3.831518181277153,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 16094
    },
    {
      "epoch": 3.8317562340058324,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 16095
    },
    {
      "epoch": 3.8319942867345116,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4488,
      "step": 16096
    },
    {
      "epoch": 3.832232339463191,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 16097
    },
    {
      "epoch": 3.8324703921918704,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 16098
    },
    {
      "epoch": 3.83270844492055,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 16099
    },
    {
      "epoch": 3.832946497649229,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 16100
    },
    {
      "epoch": 3.832946497649229,
      "eval_runtime": 0.5276,
      "eval_samples_per_second": 1.895,
      "eval_steps_per_second": 1.895,
      "step": 16100
    },
    {
      "epoch": 3.8331845503779087,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 16101
    },
    {
      "epoch": 3.8334226031065883,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3074,
      "step": 16102
    },
    {
      "epoch": 3.8336606558352675,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 16103
    },
    {
      "epoch": 3.8338987085639467,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.411,
      "step": 16104
    },
    {
      "epoch": 3.8341367612926263,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 16105
    },
    {
      "epoch": 3.834374814021306,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 16106
    },
    {
      "epoch": 3.834612866749985,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 16107
    },
    {
      "epoch": 3.8348509194786646,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 16108
    },
    {
      "epoch": 3.835088972207344,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 16109
    },
    {
      "epoch": 3.8353270249360234,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5589,
      "step": 16110
    },
    {
      "epoch": 3.835565077664703,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 16111
    },
    {
      "epoch": 3.835803130393382,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3727,
      "step": 16112
    },
    {
      "epoch": 3.8360411831220613,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 16113
    },
    {
      "epoch": 3.836279235850741,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 16114
    },
    {
      "epoch": 3.8365172885794205,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 16115
    },
    {
      "epoch": 3.8367553413080997,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 16116
    },
    {
      "epoch": 3.8369933940367793,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 16117
    },
    {
      "epoch": 3.8372314467654585,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 16118
    },
    {
      "epoch": 3.837469499494138,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2558,
      "step": 16119
    },
    {
      "epoch": 3.8377075522228172,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 16120
    },
    {
      "epoch": 3.837945604951497,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 16121
    },
    {
      "epoch": 3.838183657680176,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 16122
    },
    {
      "epoch": 3.8384217104088556,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3981,
      "step": 16123
    },
    {
      "epoch": 3.838659763137535,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 16124
    },
    {
      "epoch": 3.8388978158662144,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 16125
    },
    {
      "epoch": 3.8391358685948935,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 16126
    },
    {
      "epoch": 3.839373921323573,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 16127
    },
    {
      "epoch": 3.8396119740522527,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 16128
    },
    {
      "epoch": 3.839850026780932,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.4687,
      "step": 16129
    },
    {
      "epoch": 3.8400880795096115,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 16130
    },
    {
      "epoch": 3.8403261322382907,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 16131
    },
    {
      "epoch": 3.8405641849669703,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 16132
    },
    {
      "epoch": 3.8408022376956494,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 16133
    },
    {
      "epoch": 3.841040290424329,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 16134
    },
    {
      "epoch": 3.841278343153008,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 16135
    },
    {
      "epoch": 3.841516395881688,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 16136
    },
    {
      "epoch": 3.8417544486103674,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 16137
    },
    {
      "epoch": 3.8419925013390466,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 16138
    },
    {
      "epoch": 3.8422305540677257,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 16139
    },
    {
      "epoch": 3.8424686067964053,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 16140
    },
    {
      "epoch": 3.842706659525085,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 16141
    },
    {
      "epoch": 3.842944712253764,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 16142
    },
    {
      "epoch": 3.8431827649824437,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 16143
    },
    {
      "epoch": 3.843420817711123,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4079,
      "step": 16144
    },
    {
      "epoch": 3.8436588704398025,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 16145
    },
    {
      "epoch": 3.843896923168482,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 16146
    },
    {
      "epoch": 3.8441349758971612,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 16147
    },
    {
      "epoch": 3.8443730286258404,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 16148
    },
    {
      "epoch": 3.84461108135452,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 16149
    },
    {
      "epoch": 3.8448491340831996,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 16150
    },
    {
      "epoch": 3.8450871868118788,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 16151
    },
    {
      "epoch": 3.8453252395405584,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 16152
    },
    {
      "epoch": 3.8455632922692375,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 16153
    },
    {
      "epoch": 3.845801344997917,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 16154
    },
    {
      "epoch": 3.8460393977265963,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 16155
    },
    {
      "epoch": 3.846277450455276,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 16156
    },
    {
      "epoch": 3.846515503183955,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 16157
    },
    {
      "epoch": 3.8467535559126347,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 16158
    },
    {
      "epoch": 3.8469916086413143,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 16159
    },
    {
      "epoch": 3.8472296613699934,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 16160
    },
    {
      "epoch": 3.8474677140986726,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 16161
    },
    {
      "epoch": 3.847705766827352,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 16162
    },
    {
      "epoch": 3.847943819556032,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 16163
    },
    {
      "epoch": 3.848181872284711,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 16164
    },
    {
      "epoch": 3.8484199250133906,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 16165
    },
    {
      "epoch": 3.8486579777420697,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 16166
    },
    {
      "epoch": 3.8488960304707494,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 16167
    },
    {
      "epoch": 3.8491340831994285,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 16168
    },
    {
      "epoch": 3.849372135928108,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.399,
      "step": 16169
    },
    {
      "epoch": 3.8496101886567873,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 16170
    },
    {
      "epoch": 3.849848241385467,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 16171
    },
    {
      "epoch": 3.8500862941141465,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3349,
      "step": 16172
    },
    {
      "epoch": 3.8503243468428257,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3909,
      "step": 16173
    },
    {
      "epoch": 3.850562399571505,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 16174
    },
    {
      "epoch": 3.8508004523001844,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 16175
    },
    {
      "epoch": 3.851038505028864,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 16176
    },
    {
      "epoch": 3.851276557757543,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 16177
    },
    {
      "epoch": 3.851514610486223,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 16178
    },
    {
      "epoch": 3.851752663214902,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 16179
    },
    {
      "epoch": 3.8519907159435816,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 16180
    },
    {
      "epoch": 3.852228768672261,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 16181
    },
    {
      "epoch": 3.8524668214009403,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 16182
    },
    {
      "epoch": 3.8527048741296195,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 16183
    },
    {
      "epoch": 3.852942926858299,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 16184
    },
    {
      "epoch": 3.8531809795869787,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.341,
      "step": 16185
    },
    {
      "epoch": 3.853419032315658,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 16186
    },
    {
      "epoch": 3.8536570850443375,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1561,
      "step": 16187
    },
    {
      "epoch": 3.8538951377730166,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 16188
    },
    {
      "epoch": 3.8541331905016962,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 16189
    },
    {
      "epoch": 3.8543712432303754,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 16190
    },
    {
      "epoch": 3.854609295959055,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 16191
    },
    {
      "epoch": 3.854847348687734,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 16192
    },
    {
      "epoch": 3.8550854014164138,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3788,
      "step": 16193
    },
    {
      "epoch": 3.8553234541450934,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 16194
    },
    {
      "epoch": 3.8555615068737725,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 16195
    },
    {
      "epoch": 3.8557995596024517,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 16196
    },
    {
      "epoch": 3.8560376123311313,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 16197
    },
    {
      "epoch": 3.856275665059811,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 16198
    },
    {
      "epoch": 3.85651371778849,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 16199
    },
    {
      "epoch": 3.8567517705171697,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 16200
    },
    {
      "epoch": 3.8567517705171697,
      "eval_runtime": 0.5215,
      "eval_samples_per_second": 1.918,
      "eval_steps_per_second": 1.918,
      "step": 16200
    },
    {
      "epoch": 3.856989823245849,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5609,
      "step": 16201
    },
    {
      "epoch": 3.8572278759745284,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 16202
    },
    {
      "epoch": 3.857465928703208,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 16203
    },
    {
      "epoch": 3.857703981431887,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 16204
    },
    {
      "epoch": 3.8579420341605664,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 16205
    },
    {
      "epoch": 3.858180086889246,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 16206
    },
    {
      "epoch": 3.8584181396179256,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 16207
    },
    {
      "epoch": 3.8586561923466047,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 16208
    },
    {
      "epoch": 3.8588942450752843,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 16209
    },
    {
      "epoch": 3.8591322978039635,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 16210
    },
    {
      "epoch": 3.859370350532643,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2781,
      "step": 16211
    },
    {
      "epoch": 3.8596084032613223,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 16212
    },
    {
      "epoch": 3.859846455990002,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 16213
    },
    {
      "epoch": 3.860084508718681,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 16214
    },
    {
      "epoch": 3.8603225614473606,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 16215
    },
    {
      "epoch": 3.8605606141760402,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 16216
    },
    {
      "epoch": 3.8607986669047194,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 16217
    },
    {
      "epoch": 3.8610367196333986,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 16218
    },
    {
      "epoch": 3.861274772362078,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 16219
    },
    {
      "epoch": 3.8615128250907578,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 16220
    },
    {
      "epoch": 3.861750877819437,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 16221
    },
    {
      "epoch": 3.8619889305481165,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 16222
    },
    {
      "epoch": 3.8622269832767957,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 16223
    },
    {
      "epoch": 3.8624650360054753,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 16224
    },
    {
      "epoch": 3.8627030887341545,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 16225
    },
    {
      "epoch": 3.862941141462834,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 16226
    },
    {
      "epoch": 3.8631791941915132,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 16227
    },
    {
      "epoch": 3.863417246920193,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 16228
    },
    {
      "epoch": 3.8636552996488724,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 16229
    },
    {
      "epoch": 3.8638933523775516,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 16230
    },
    {
      "epoch": 3.8641314051062308,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 16231
    },
    {
      "epoch": 3.8643694578349104,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 16232
    },
    {
      "epoch": 3.86460751056359,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 16233
    },
    {
      "epoch": 3.864845563292269,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 16234
    },
    {
      "epoch": 3.8650836160209487,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 16235
    },
    {
      "epoch": 3.865321668749628,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.5133,
      "step": 16236
    },
    {
      "epoch": 3.8655597214783075,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 16237
    },
    {
      "epoch": 3.865797774206987,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4233,
      "step": 16238
    },
    {
      "epoch": 3.8660358269356663,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 16239
    },
    {
      "epoch": 3.8662738796643454,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 16240
    },
    {
      "epoch": 3.866511932393025,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3273,
      "step": 16241
    },
    {
      "epoch": 3.8667499851217046,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 16242
    },
    {
      "epoch": 3.866988037850384,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.361,
      "step": 16243
    },
    {
      "epoch": 3.8672260905790634,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 16244
    },
    {
      "epoch": 3.8674641433077426,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 16245
    },
    {
      "epoch": 3.867702196036422,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 16246
    },
    {
      "epoch": 3.8679402487651013,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5071,
      "step": 16247
    },
    {
      "epoch": 3.868178301493781,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 16248
    },
    {
      "epoch": 3.86841635422246,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 16249
    },
    {
      "epoch": 3.8686544069511397,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 16250
    },
    {
      "epoch": 3.8688924596798193,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 16251
    },
    {
      "epoch": 3.8691305124084985,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 16252
    },
    {
      "epoch": 3.8693685651371776,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 16253
    },
    {
      "epoch": 3.8696066178658572,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 16254
    },
    {
      "epoch": 3.869844670594537,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 16255
    },
    {
      "epoch": 3.870082723323216,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 16256
    },
    {
      "epoch": 3.8703207760518956,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3876,
      "step": 16257
    },
    {
      "epoch": 3.870558828780575,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 16258
    },
    {
      "epoch": 3.8707968815092544,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 16259
    },
    {
      "epoch": 3.8710349342379335,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 16260
    },
    {
      "epoch": 3.871272986966613,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 16261
    },
    {
      "epoch": 3.8715110396952923,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 16262
    },
    {
      "epoch": 3.871749092423972,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 16263
    },
    {
      "epoch": 3.8719871451526515,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 16264
    },
    {
      "epoch": 3.8722251978813307,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 16265
    },
    {
      "epoch": 3.87246325061001,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 16266
    },
    {
      "epoch": 3.8727013033386894,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 16267
    },
    {
      "epoch": 3.872939356067369,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 16268
    },
    {
      "epoch": 3.873177408796048,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 16269
    },
    {
      "epoch": 3.873415461524728,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3197,
      "step": 16270
    },
    {
      "epoch": 3.873653514253407,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 16271
    },
    {
      "epoch": 3.8738915669820866,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 16272
    },
    {
      "epoch": 3.874129619710766,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 16273
    },
    {
      "epoch": 3.8743676724394454,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 16274
    },
    {
      "epoch": 3.8746057251681245,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 16275
    },
    {
      "epoch": 3.874843777896804,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 16276
    },
    {
      "epoch": 3.8750818306254837,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 16277
    },
    {
      "epoch": 3.875319883354163,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9363,
      "step": 16278
    },
    {
      "epoch": 3.8755579360828425,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 16279
    },
    {
      "epoch": 3.8757959888115217,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.365,
      "step": 16280
    },
    {
      "epoch": 3.8760340415402013,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 16281
    },
    {
      "epoch": 3.8762720942688804,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 16282
    },
    {
      "epoch": 3.87651014699756,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 16283
    },
    {
      "epoch": 3.876748199726239,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 16284
    },
    {
      "epoch": 3.876986252454919,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 16285
    },
    {
      "epoch": 3.8772243051835984,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3336,
      "step": 16286
    },
    {
      "epoch": 3.8774623579122776,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 16287
    },
    {
      "epoch": 3.8777004106409567,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 16288
    },
    {
      "epoch": 3.8779384633696363,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4302,
      "step": 16289
    },
    {
      "epoch": 3.878176516098316,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 16290
    },
    {
      "epoch": 3.878414568826995,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 16291
    },
    {
      "epoch": 3.8786526215556747,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 16292
    },
    {
      "epoch": 3.878890674284354,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 16293
    },
    {
      "epoch": 3.8791287270130335,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.293,
      "step": 16294
    },
    {
      "epoch": 3.8793667797417126,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 16295
    },
    {
      "epoch": 3.8796048324703922,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 16296
    },
    {
      "epoch": 3.8798428851990714,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 16297
    },
    {
      "epoch": 3.880080937927751,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 16298
    },
    {
      "epoch": 3.8803189906564306,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 16299
    },
    {
      "epoch": 3.8805570433851098,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 16300
    },
    {
      "epoch": 3.8805570433851098,
      "eval_runtime": 0.5203,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 16300
    },
    {
      "epoch": 3.8807950961137894,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1574,
      "step": 16301
    },
    {
      "epoch": 3.8810331488424685,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 16302
    },
    {
      "epoch": 3.881271201571148,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 16303
    },
    {
      "epoch": 3.8815092542998273,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 16304
    },
    {
      "epoch": 3.881747307028507,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 16305
    },
    {
      "epoch": 3.881985359757186,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 16306
    },
    {
      "epoch": 3.8822234124858657,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 16307
    },
    {
      "epoch": 3.8824614652145453,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4761,
      "step": 16308
    },
    {
      "epoch": 3.8826995179432244,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 16309
    },
    {
      "epoch": 3.8829375706719036,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 16310
    },
    {
      "epoch": 3.883175623400583,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 16311
    },
    {
      "epoch": 3.883413676129263,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3366,
      "step": 16312
    },
    {
      "epoch": 3.883651728857942,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 16313
    },
    {
      "epoch": 3.8838897815866216,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 16314
    },
    {
      "epoch": 3.8841278343153007,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 16315
    },
    {
      "epoch": 3.8843658870439803,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 16316
    },
    {
      "epoch": 3.8846039397726595,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 16317
    },
    {
      "epoch": 3.884841992501339,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 16318
    },
    {
      "epoch": 3.8850800452300183,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 16319
    },
    {
      "epoch": 3.885318097958698,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 16320
    },
    {
      "epoch": 3.8855561506873775,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 16321
    },
    {
      "epoch": 3.8857942034160566,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 16322
    },
    {
      "epoch": 3.886032256144736,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 16323
    },
    {
      "epoch": 3.8862703088734154,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 16324
    },
    {
      "epoch": 3.886508361602095,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8273,
      "step": 16325
    },
    {
      "epoch": 3.886746414330774,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 16326
    },
    {
      "epoch": 3.8869844670594538,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 16327
    },
    {
      "epoch": 3.887222519788133,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3602,
      "step": 16328
    },
    {
      "epoch": 3.8874605725168125,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 16329
    },
    {
      "epoch": 3.887698625245492,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 16330
    },
    {
      "epoch": 3.8879366779741713,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 16331
    },
    {
      "epoch": 3.8881747307028505,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1629,
      "step": 16332
    },
    {
      "epoch": 3.88841278343153,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 16333
    },
    {
      "epoch": 3.8886508361602097,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 16334
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.285,
      "step": 16335
    },
    {
      "epoch": 3.8891269416175684,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 16336
    },
    {
      "epoch": 3.8893649943462476,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 16337
    },
    {
      "epoch": 3.889603047074927,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 16338
    },
    {
      "epoch": 3.8898410998036064,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 16339
    },
    {
      "epoch": 3.890079152532286,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 16340
    },
    {
      "epoch": 3.890317205260965,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.5322,
      "step": 16341
    },
    {
      "epoch": 3.8905552579896447,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 16342
    },
    {
      "epoch": 3.8907933107183243,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 16343
    },
    {
      "epoch": 3.8910313634470035,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 16344
    },
    {
      "epoch": 3.8912694161756827,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3076,
      "step": 16345
    },
    {
      "epoch": 3.8915074689043623,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 16346
    },
    {
      "epoch": 3.891745521633042,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 16347
    },
    {
      "epoch": 3.891983574361721,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4821,
      "step": 16348
    },
    {
      "epoch": 3.8922216270904006,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 16349
    },
    {
      "epoch": 3.89245967981908,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 16350
    },
    {
      "epoch": 3.8926977325477594,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 16351
    },
    {
      "epoch": 3.8929357852764386,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 16352
    },
    {
      "epoch": 3.893173838005118,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2816,
      "step": 16353
    },
    {
      "epoch": 3.8934118907337973,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 16354
    },
    {
      "epoch": 3.893649943462477,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 16355
    },
    {
      "epoch": 3.8938879961911566,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 16356
    },
    {
      "epoch": 3.8941260489198357,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4548,
      "step": 16357
    },
    {
      "epoch": 3.894364101648515,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 16358
    },
    {
      "epoch": 3.8946021543771945,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2532,
      "step": 16359
    },
    {
      "epoch": 3.894840207105874,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 16360
    },
    {
      "epoch": 3.8950782598345532,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 16361
    },
    {
      "epoch": 3.895316312563233,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 16362
    },
    {
      "epoch": 3.895554365291912,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 16363
    },
    {
      "epoch": 3.8957924180205916,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 16364
    },
    {
      "epoch": 3.8960304707492712,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 16365
    },
    {
      "epoch": 3.8962685234779504,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 16366
    },
    {
      "epoch": 3.8965065762066295,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 16367
    },
    {
      "epoch": 3.896744628935309,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 16368
    },
    {
      "epoch": 3.8969826816639888,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 16369
    },
    {
      "epoch": 3.897220734392668,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 16370
    },
    {
      "epoch": 3.8974587871213475,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 16371
    },
    {
      "epoch": 3.8976968398500267,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 16372
    },
    {
      "epoch": 3.8979348925787063,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 16373
    },
    {
      "epoch": 3.8981729453073855,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3972,
      "step": 16374
    },
    {
      "epoch": 3.898410998036065,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2979,
      "step": 16375
    },
    {
      "epoch": 3.898649050764744,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.9096,
      "step": 16376
    },
    {
      "epoch": 3.898887103493424,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2851,
      "step": 16377
    },
    {
      "epoch": 3.8991251562221034,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 16378
    },
    {
      "epoch": 3.8993632089507826,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3496,
      "step": 16379
    },
    {
      "epoch": 3.8996012616794618,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 16380
    },
    {
      "epoch": 3.8998393144081414,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 16381
    },
    {
      "epoch": 3.900077367136821,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 16382
    },
    {
      "epoch": 3.9003154198655,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 16383
    },
    {
      "epoch": 3.9005534725941797,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 16384
    },
    {
      "epoch": 3.900791525322859,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 16385
    },
    {
      "epoch": 3.9010295780515385,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 16386
    },
    {
      "epoch": 3.9012676307802177,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 16387
    },
    {
      "epoch": 3.9015056835088973,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 16388
    },
    {
      "epoch": 3.9017437362375764,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 16389
    },
    {
      "epoch": 3.901981788966256,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 16390
    },
    {
      "epoch": 3.9022198416949356,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 16391
    },
    {
      "epoch": 3.902457894423615,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 16392
    },
    {
      "epoch": 3.9026959471522944,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 16393
    },
    {
      "epoch": 3.9029339998809736,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 16394
    },
    {
      "epoch": 3.903172052609653,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 16395
    },
    {
      "epoch": 3.9034101053383323,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 16396
    },
    {
      "epoch": 3.903648158067012,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.367,
      "step": 16397
    },
    {
      "epoch": 3.903886210795691,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 16398
    },
    {
      "epoch": 3.9041242635243707,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 16399
    },
    {
      "epoch": 3.9043623162530503,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 16400
    },
    {
      "epoch": 3.9043623162530503,
      "eval_runtime": 0.5127,
      "eval_samples_per_second": 1.95,
      "eval_steps_per_second": 1.95,
      "step": 16400
    },
    {
      "epoch": 3.9046003689817295,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 16401
    },
    {
      "epoch": 3.9048384217104086,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 16402
    },
    {
      "epoch": 3.9050764744390882,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 16403
    },
    {
      "epoch": 3.905314527167768,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4868,
      "step": 16404
    },
    {
      "epoch": 3.905552579896447,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4235,
      "step": 16405
    },
    {
      "epoch": 3.9057906326251266,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 16406
    },
    {
      "epoch": 3.9060286853538058,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 16407
    },
    {
      "epoch": 3.9062667380824854,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2789,
      "step": 16408
    },
    {
      "epoch": 3.9065047908111645,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 16409
    },
    {
      "epoch": 3.906742843539844,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 16410
    },
    {
      "epoch": 3.9069808962685233,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 16411
    },
    {
      "epoch": 3.907218948997203,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 16412
    },
    {
      "epoch": 3.9074570017258825,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 16413
    },
    {
      "epoch": 3.9076950544545617,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 16414
    },
    {
      "epoch": 3.907933107183241,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 16415
    },
    {
      "epoch": 3.9081711599119204,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 16416
    },
    {
      "epoch": 3.9084092126406,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 16417
    },
    {
      "epoch": 3.908647265369279,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 16418
    },
    {
      "epoch": 3.908885318097959,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 16419
    },
    {
      "epoch": 3.909123370826638,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3193,
      "step": 16420
    },
    {
      "epoch": 3.9093614235553176,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 16421
    },
    {
      "epoch": 3.909599476283997,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 16422
    },
    {
      "epoch": 3.9098375290126763,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 16423
    },
    {
      "epoch": 3.9100755817413555,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3874,
      "step": 16424
    },
    {
      "epoch": 3.910313634470035,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 16425
    },
    {
      "epoch": 3.9105516871987147,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 16426
    },
    {
      "epoch": 3.910789739927394,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 16427
    },
    {
      "epoch": 3.9110277926560735,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 16428
    },
    {
      "epoch": 3.9112658453847526,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.5697,
      "step": 16429
    },
    {
      "epoch": 3.9115038981134322,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 16430
    },
    {
      "epoch": 3.9117419508421114,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 16431
    },
    {
      "epoch": 3.911980003570791,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 16432
    },
    {
      "epoch": 3.91221805629947,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 16433
    },
    {
      "epoch": 3.9124561090281498,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 16434
    },
    {
      "epoch": 3.9126941617568294,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3827,
      "step": 16435
    },
    {
      "epoch": 3.9129322144855085,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 16436
    },
    {
      "epoch": 3.9131702672141877,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 16437
    },
    {
      "epoch": 3.9134083199428673,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 16438
    },
    {
      "epoch": 3.913646372671547,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 16439
    },
    {
      "epoch": 3.913884425400226,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 16440
    },
    {
      "epoch": 3.9141224781289057,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 16441
    },
    {
      "epoch": 3.914360530857585,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4933,
      "step": 16442
    },
    {
      "epoch": 3.9145985835862644,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 16443
    },
    {
      "epoch": 3.9148366363149436,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 16444
    },
    {
      "epoch": 3.915074689043623,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3695,
      "step": 16445
    },
    {
      "epoch": 3.9153127417723024,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 16446
    },
    {
      "epoch": 3.915550794500982,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 16447
    },
    {
      "epoch": 3.9157888472296616,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 16448
    },
    {
      "epoch": 3.9160268999583407,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2332,
      "step": 16449
    },
    {
      "epoch": 3.91626495268702,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 16450
    },
    {
      "epoch": 3.9165030054156995,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 16451
    },
    {
      "epoch": 3.916741058144379,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 16452
    },
    {
      "epoch": 3.9169791108730583,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 16453
    },
    {
      "epoch": 3.917217163601738,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 16454
    },
    {
      "epoch": 3.917455216330417,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 16455
    },
    {
      "epoch": 3.9176932690590967,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 16456
    },
    {
      "epoch": 3.9179313217877763,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 16457
    },
    {
      "epoch": 3.9181693745164554,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 16458
    },
    {
      "epoch": 3.9184074272451346,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 16459
    },
    {
      "epoch": 3.918645479973814,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 16460
    },
    {
      "epoch": 3.918883532702494,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 16461
    },
    {
      "epoch": 3.919121585431173,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 16462
    },
    {
      "epoch": 3.9193596381598526,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 16463
    },
    {
      "epoch": 3.9195976908885317,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 16464
    },
    {
      "epoch": 3.9198357436172113,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 16465
    },
    {
      "epoch": 3.9200737963458905,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4208,
      "step": 16466
    },
    {
      "epoch": 3.92031184907457,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 16467
    },
    {
      "epoch": 3.9205499018032492,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 16468
    },
    {
      "epoch": 3.920787954531929,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 16469
    },
    {
      "epoch": 3.9210260072606085,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2542,
      "step": 16470
    },
    {
      "epoch": 3.9212640599892876,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 16471
    },
    {
      "epoch": 3.921502112717967,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3237,
      "step": 16472
    },
    {
      "epoch": 3.9217401654466464,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 16473
    },
    {
      "epoch": 3.921978218175326,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 16474
    },
    {
      "epoch": 3.922216270904005,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 16475
    },
    {
      "epoch": 3.9224543236326848,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 16476
    },
    {
      "epoch": 3.922692376361364,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 16477
    },
    {
      "epoch": 3.9229304290900435,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 16478
    },
    {
      "epoch": 3.9231684818187227,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 16479
    },
    {
      "epoch": 3.9234065345474023,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 16480
    },
    {
      "epoch": 3.9236445872760815,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3533,
      "step": 16481
    },
    {
      "epoch": 3.923882640004761,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 16482
    },
    {
      "epoch": 3.9241206927334407,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 16483
    },
    {
      "epoch": 3.92435874546212,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 16484
    },
    {
      "epoch": 3.924596798190799,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 16485
    },
    {
      "epoch": 3.9248348509194786,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 16486
    },
    {
      "epoch": 3.925072903648158,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2578,
      "step": 16487
    },
    {
      "epoch": 3.9253109563768374,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 16488
    },
    {
      "epoch": 3.925549009105517,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 16489
    },
    {
      "epoch": 3.925787061834196,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 16490
    },
    {
      "epoch": 3.9260251145628757,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4501,
      "step": 16491
    },
    {
      "epoch": 3.9262631672915553,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 16492
    },
    {
      "epoch": 3.9265012200202345,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 16493
    },
    {
      "epoch": 3.9267392727489137,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 16494
    },
    {
      "epoch": 3.9269773254775933,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 16495
    },
    {
      "epoch": 3.927215378206273,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 16496
    },
    {
      "epoch": 3.927453430934952,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 16497
    },
    {
      "epoch": 3.9276914836636316,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3437,
      "step": 16498
    },
    {
      "epoch": 3.927929536392311,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 16499
    },
    {
      "epoch": 3.9281675891209904,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 16500
    },
    {
      "epoch": 3.9281675891209904,
      "eval_runtime": 0.5241,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 16500
    },
    {
      "epoch": 3.9284056418496696,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 16501
    },
    {
      "epoch": 3.928643694578349,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 16502
    },
    {
      "epoch": 3.9288817473070283,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3075,
      "step": 16503
    },
    {
      "epoch": 3.929119800035708,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 16504
    },
    {
      "epoch": 3.9293578527643875,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 16505
    },
    {
      "epoch": 3.9295959054930667,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 16506
    },
    {
      "epoch": 3.929833958221746,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 16507
    },
    {
      "epoch": 3.9300720109504255,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.4691,
      "step": 16508
    },
    {
      "epoch": 3.930310063679105,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 16509
    },
    {
      "epoch": 3.9305481164077842,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 16510
    },
    {
      "epoch": 3.930786169136464,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3071,
      "step": 16511
    },
    {
      "epoch": 3.931024221865143,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 16512
    },
    {
      "epoch": 3.9312622745938226,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 16513
    },
    {
      "epoch": 3.931500327322502,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4685,
      "step": 16514
    },
    {
      "epoch": 3.9317383800511814,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 16515
    },
    {
      "epoch": 3.9319764327798605,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 16516
    },
    {
      "epoch": 3.93221448550854,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 16517
    },
    {
      "epoch": 3.9324525382372197,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 16518
    },
    {
      "epoch": 3.932690590965899,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 16519
    },
    {
      "epoch": 3.9329286436945785,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 16520
    },
    {
      "epoch": 3.9331666964232577,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 16521
    },
    {
      "epoch": 3.9334047491519373,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 16522
    },
    {
      "epoch": 3.9336428018806164,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4165,
      "step": 16523
    },
    {
      "epoch": 3.933880854609296,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 16524
    },
    {
      "epoch": 3.934118907337975,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 16525
    },
    {
      "epoch": 3.934356960066655,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 16526
    },
    {
      "epoch": 3.9345950127953344,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 16527
    },
    {
      "epoch": 3.9348330655240136,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 16528
    },
    {
      "epoch": 3.9350711182526927,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 16529
    },
    {
      "epoch": 3.9353091709813723,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4634,
      "step": 16530
    },
    {
      "epoch": 3.935547223710052,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1948,
      "step": 16531
    },
    {
      "epoch": 3.935785276438731,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.392,
      "step": 16532
    },
    {
      "epoch": 3.9360233291674107,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 16533
    },
    {
      "epoch": 3.93626138189609,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 16534
    },
    {
      "epoch": 3.9364994346247695,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 16535
    },
    {
      "epoch": 3.9367374873534486,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 16536
    },
    {
      "epoch": 3.9369755400821282,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 16537
    },
    {
      "epoch": 3.9372135928108074,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 16538
    },
    {
      "epoch": 3.937451645539487,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2786,
      "step": 16539
    },
    {
      "epoch": 3.9376896982681666,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4929,
      "step": 16540
    },
    {
      "epoch": 3.9379277509968458,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 16541
    },
    {
      "epoch": 3.938165803725525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 16542
    },
    {
      "epoch": 3.9384038564542045,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 16543
    },
    {
      "epoch": 3.938641909182884,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 16544
    },
    {
      "epoch": 3.9388799619115633,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 16545
    },
    {
      "epoch": 3.939118014640243,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7062,
      "step": 16546
    },
    {
      "epoch": 3.939356067368922,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 16547
    },
    {
      "epoch": 3.9395941200976017,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3037,
      "step": 16548
    },
    {
      "epoch": 3.9398321728262813,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.5648,
      "step": 16549
    },
    {
      "epoch": 3.9400702255549604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 16550
    },
    {
      "epoch": 3.9403082782836396,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 16551
    },
    {
      "epoch": 3.940546331012319,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 16552
    },
    {
      "epoch": 3.940784383740999,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 16553
    },
    {
      "epoch": 3.941022436469678,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 16554
    },
    {
      "epoch": 3.9412604891983576,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4566,
      "step": 16555
    },
    {
      "epoch": 3.9414985419270367,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 16556
    },
    {
      "epoch": 3.9417365946557164,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 16557
    },
    {
      "epoch": 3.9419746473843955,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 16558
    },
    {
      "epoch": 3.942212700113075,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 16559
    },
    {
      "epoch": 3.9424507528417543,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 16560
    },
    {
      "epoch": 3.942688805570434,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 16561
    },
    {
      "epoch": 3.9429268582991135,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 16562
    },
    {
      "epoch": 3.9431649110277927,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 16563
    },
    {
      "epoch": 3.943402963756472,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 16564
    },
    {
      "epoch": 3.9436410164851514,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 16565
    },
    {
      "epoch": 3.943879069213831,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 16566
    },
    {
      "epoch": 3.94411712194251,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 16567
    },
    {
      "epoch": 3.94435517467119,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2559,
      "step": 16568
    },
    {
      "epoch": 3.944593227399869,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 16569
    },
    {
      "epoch": 3.9448312801285486,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 16570
    },
    {
      "epoch": 3.9450693328572277,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2937,
      "step": 16571
    },
    {
      "epoch": 3.9453073855859073,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.4166,
      "step": 16572
    },
    {
      "epoch": 3.9455454383145865,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 16573
    },
    {
      "epoch": 3.945783491043266,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2756,
      "step": 16574
    },
    {
      "epoch": 3.9460215437719457,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 16575
    },
    {
      "epoch": 3.946259596500625,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 16576
    },
    {
      "epoch": 3.946497649229304,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 16577
    },
    {
      "epoch": 3.9467357019579836,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 16578
    },
    {
      "epoch": 3.9469737546866632,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 16579
    },
    {
      "epoch": 3.9472118074153424,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 16580
    },
    {
      "epoch": 3.947449860144022,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 16581
    },
    {
      "epoch": 3.947687912872701,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 16582
    },
    {
      "epoch": 3.9479259656013808,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 16583
    },
    {
      "epoch": 3.9481640183300604,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3483,
      "step": 16584
    },
    {
      "epoch": 3.9484020710587395,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 16585
    },
    {
      "epoch": 3.9486401237874187,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 16586
    },
    {
      "epoch": 3.9488781765160983,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 16587
    },
    {
      "epoch": 3.949116229244778,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4745,
      "step": 16588
    },
    {
      "epoch": 3.949354281973457,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 16589
    },
    {
      "epoch": 3.9495923347021367,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 16590
    },
    {
      "epoch": 3.949830387430816,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 16591
    },
    {
      "epoch": 3.9500684401594954,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2587,
      "step": 16592
    },
    {
      "epoch": 3.9503064928881746,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 16593
    },
    {
      "epoch": 3.950544545616854,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 16594
    },
    {
      "epoch": 3.9507825983455334,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 16595
    },
    {
      "epoch": 3.951020651074213,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 16596
    },
    {
      "epoch": 3.9512587038028926,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5603,
      "step": 16597
    },
    {
      "epoch": 3.9514967565315717,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 16598
    },
    {
      "epoch": 3.951734809260251,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 16599
    },
    {
      "epoch": 3.9519728619889305,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 16600
    },
    {
      "epoch": 3.9519728619889305,
      "eval_runtime": 0.5226,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 16600
    },
    {
      "epoch": 3.95221091471761,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 16601
    },
    {
      "epoch": 3.9524489674462893,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 16602
    },
    {
      "epoch": 3.952687020174969,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 16603
    },
    {
      "epoch": 3.952925072903648,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 16604
    },
    {
      "epoch": 3.9531631256323276,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 16605
    },
    {
      "epoch": 3.9534011783610072,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4984,
      "step": 16606
    },
    {
      "epoch": 3.9536392310896864,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 16607
    },
    {
      "epoch": 3.9538772838183656,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2812,
      "step": 16608
    },
    {
      "epoch": 3.954115336547045,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 16609
    },
    {
      "epoch": 3.9543533892757248,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 16610
    },
    {
      "epoch": 3.954591442004404,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 16611
    },
    {
      "epoch": 3.9548294947330835,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 16612
    },
    {
      "epoch": 3.9550675474617627,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 16613
    },
    {
      "epoch": 3.9553056001904423,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 16614
    },
    {
      "epoch": 3.9555436529191215,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3626,
      "step": 16615
    },
    {
      "epoch": 3.955781705647801,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 16616
    },
    {
      "epoch": 3.9560197583764802,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 16617
    },
    {
      "epoch": 3.95625781110516,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 16618
    },
    {
      "epoch": 3.9564958638338394,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 16619
    },
    {
      "epoch": 3.9567339165625186,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 16620
    },
    {
      "epoch": 3.9569719692911978,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 16621
    },
    {
      "epoch": 3.9572100220198774,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 16622
    },
    {
      "epoch": 3.957448074748557,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 16623
    },
    {
      "epoch": 3.957686127477236,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 16624
    },
    {
      "epoch": 3.9579241802059157,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 16625
    },
    {
      "epoch": 3.958162232934595,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 16626
    },
    {
      "epoch": 3.9584002856632745,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 16627
    },
    {
      "epoch": 3.9586383383919537,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 16628
    },
    {
      "epoch": 3.9588763911206333,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 16629
    },
    {
      "epoch": 3.9591144438493124,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 16630
    },
    {
      "epoch": 3.959352496577992,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 16631
    },
    {
      "epoch": 3.9595905493066716,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 16632
    },
    {
      "epoch": 3.959828602035351,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 16633
    },
    {
      "epoch": 3.96006665476403,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 16634
    },
    {
      "epoch": 3.9603047074927096,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 16635
    },
    {
      "epoch": 3.960542760221389,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 16636
    },
    {
      "epoch": 3.9607808129500683,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2577,
      "step": 16637
    },
    {
      "epoch": 3.961018865678748,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.5716,
      "step": 16638
    },
    {
      "epoch": 3.961256918407427,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 16639
    },
    {
      "epoch": 3.9614949711361067,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 16640
    },
    {
      "epoch": 3.9617330238647863,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 16641
    },
    {
      "epoch": 3.9619710765934655,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2549,
      "step": 16642
    },
    {
      "epoch": 3.9622091293221446,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 16643
    },
    {
      "epoch": 3.9624471820508242,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 16644
    },
    {
      "epoch": 3.962685234779504,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4795,
      "step": 16645
    },
    {
      "epoch": 3.962923287508183,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 16646
    },
    {
      "epoch": 3.9631613402368626,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 16647
    },
    {
      "epoch": 3.963399392965542,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3606,
      "step": 16648
    },
    {
      "epoch": 3.9636374456942214,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 16649
    },
    {
      "epoch": 3.9638754984229005,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.203,
      "step": 16650
    },
    {
      "epoch": 3.96411355115158,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 16651
    },
    {
      "epoch": 3.9643516038802593,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 16652
    },
    {
      "epoch": 3.964589656608939,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 16653
    },
    {
      "epoch": 3.9648277093376185,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 16654
    },
    {
      "epoch": 3.9650657620662977,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 16655
    },
    {
      "epoch": 3.965303814794977,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 16656
    },
    {
      "epoch": 3.9655418675236564,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 16657
    },
    {
      "epoch": 3.965779920252336,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 16658
    },
    {
      "epoch": 3.966017972981015,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 16659
    },
    {
      "epoch": 3.966256025709695,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 16660
    },
    {
      "epoch": 3.966494078438374,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 16661
    },
    {
      "epoch": 3.9667321311670536,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 16662
    },
    {
      "epoch": 3.9669701838957327,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 16663
    },
    {
      "epoch": 3.9672082366244124,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 16664
    },
    {
      "epoch": 3.9674462893530915,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 16665
    },
    {
      "epoch": 3.967684342081771,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 16666
    },
    {
      "epoch": 3.9679223948104507,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 16667
    },
    {
      "epoch": 3.96816044753913,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 16668
    },
    {
      "epoch": 3.968398500267809,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 16669
    },
    {
      "epoch": 3.9686365529964887,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 16670
    },
    {
      "epoch": 3.9688746057251683,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 16671
    },
    {
      "epoch": 3.9691126584538474,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 16672
    },
    {
      "epoch": 3.969350711182527,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 16673
    },
    {
      "epoch": 3.969588763911206,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 16674
    },
    {
      "epoch": 3.969826816639886,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 16675
    },
    {
      "epoch": 3.9700648693685654,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 16676
    },
    {
      "epoch": 3.9703029220972446,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 16677
    },
    {
      "epoch": 3.9705409748259237,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 16678
    },
    {
      "epoch": 3.9707790275546033,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 16679
    },
    {
      "epoch": 3.971017080283283,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 16680
    },
    {
      "epoch": 3.971255133011962,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 16681
    },
    {
      "epoch": 3.9714931857406417,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 16682
    },
    {
      "epoch": 3.971731238469321,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 16683
    },
    {
      "epoch": 3.9719692911980005,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.386,
      "step": 16684
    },
    {
      "epoch": 3.9722073439266796,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 16685
    },
    {
      "epoch": 3.9724453966553592,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 16686
    },
    {
      "epoch": 3.9726834493840384,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 16687
    },
    {
      "epoch": 3.972921502112718,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 16688
    },
    {
      "epoch": 3.9731595548413976,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 16689
    },
    {
      "epoch": 3.9733976075700768,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 16690
    },
    {
      "epoch": 3.973635660298756,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 16691
    },
    {
      "epoch": 3.9738737130274355,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 16692
    },
    {
      "epoch": 3.974111765756115,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 16693
    },
    {
      "epoch": 3.9743498184847943,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 16694
    },
    {
      "epoch": 3.974587871213474,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 16695
    },
    {
      "epoch": 3.974825923942153,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 16696
    },
    {
      "epoch": 3.9750639766708327,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 16697
    },
    {
      "epoch": 3.975302029399512,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 16698
    },
    {
      "epoch": 3.9755400821281914,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 16699
    },
    {
      "epoch": 3.9757781348568706,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 16700
    },
    {
      "epoch": 3.9757781348568706,
      "eval_runtime": 0.5293,
      "eval_samples_per_second": 1.889,
      "eval_steps_per_second": 1.889,
      "step": 16700
    },
    {
      "epoch": 3.97601618758555,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 16701
    },
    {
      "epoch": 3.97625424031423,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 16702
    },
    {
      "epoch": 3.976492293042909,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 16703
    },
    {
      "epoch": 3.9767303457715886,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 16704
    },
    {
      "epoch": 3.9769683985002677,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.5568,
      "step": 16705
    },
    {
      "epoch": 3.9772064512289473,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3526,
      "step": 16706
    },
    {
      "epoch": 3.9774445039576265,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 16707
    },
    {
      "epoch": 3.977682556686306,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 16708
    },
    {
      "epoch": 3.9779206094149853,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 16709
    },
    {
      "epoch": 3.978158662143665,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3803,
      "step": 16710
    },
    {
      "epoch": 3.9783967148723445,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 16711
    },
    {
      "epoch": 3.9786347676010236,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 16712
    },
    {
      "epoch": 3.978872820329703,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 16713
    },
    {
      "epoch": 3.9791108730583824,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 16714
    },
    {
      "epoch": 3.979348925787062,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.379,
      "step": 16715
    },
    {
      "epoch": 3.979586978515741,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 16716
    },
    {
      "epoch": 3.9798250312444208,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2785,
      "step": 16717
    },
    {
      "epoch": 3.9800630839731,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 16718
    },
    {
      "epoch": 3.9803011367017795,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 16719
    },
    {
      "epoch": 3.9805391894304587,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 16720
    },
    {
      "epoch": 3.9807772421591383,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 16721
    },
    {
      "epoch": 3.9810152948878175,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 16722
    },
    {
      "epoch": 3.981253347616497,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 16723
    },
    {
      "epoch": 3.9814914003451767,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 16724
    },
    {
      "epoch": 3.981729453073856,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 16725
    },
    {
      "epoch": 3.981967505802535,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5528,
      "step": 16726
    },
    {
      "epoch": 3.9822055585312146,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 16727
    },
    {
      "epoch": 3.982443611259894,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 16728
    },
    {
      "epoch": 3.9826816639885734,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2387,
      "step": 16729
    },
    {
      "epoch": 3.982919716717253,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 16730
    },
    {
      "epoch": 3.983157769445932,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 16731
    },
    {
      "epoch": 3.9833958221746117,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 16732
    },
    {
      "epoch": 3.9836338749032913,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 16733
    },
    {
      "epoch": 3.9838719276319705,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 16734
    },
    {
      "epoch": 3.9841099803606497,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 16735
    },
    {
      "epoch": 3.9843480330893293,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 16736
    },
    {
      "epoch": 3.984586085818009,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 16737
    },
    {
      "epoch": 3.984824138546688,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 16738
    },
    {
      "epoch": 3.9850621912753676,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 16739
    },
    {
      "epoch": 3.985300244004047,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4185,
      "step": 16740
    },
    {
      "epoch": 3.9855382967327264,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 16741
    },
    {
      "epoch": 3.9857763494614056,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.446,
      "step": 16742
    },
    {
      "epoch": 3.986014402190085,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 16743
    },
    {
      "epoch": 3.9862524549187643,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 16744
    },
    {
      "epoch": 3.986490507647444,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 16745
    },
    {
      "epoch": 3.9867285603761236,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 16746
    },
    {
      "epoch": 3.9869666131048027,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 16747
    },
    {
      "epoch": 3.987204665833482,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 16748
    },
    {
      "epoch": 3.9874427185621615,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 16749
    },
    {
      "epoch": 3.987680771290841,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 16750
    },
    {
      "epoch": 3.9879188240195202,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 16751
    },
    {
      "epoch": 3.9881568767482,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 16752
    },
    {
      "epoch": 3.988394929476879,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4235,
      "step": 16753
    },
    {
      "epoch": 3.9886329822055586,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 16754
    },
    {
      "epoch": 3.988871034934238,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 16755
    },
    {
      "epoch": 3.9891090876629174,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 16756
    },
    {
      "epoch": 3.9893471403915965,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 16757
    },
    {
      "epoch": 3.989585193120276,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3295,
      "step": 16758
    },
    {
      "epoch": 3.9898232458489558,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.531,
      "step": 16759
    },
    {
      "epoch": 3.990061298577635,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 16760
    },
    {
      "epoch": 3.990299351306314,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 16761
    },
    {
      "epoch": 3.9905374040349937,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 16762
    },
    {
      "epoch": 3.9907754567636733,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3495,
      "step": 16763
    },
    {
      "epoch": 3.9910135094923525,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 16764
    },
    {
      "epoch": 3.991251562221032,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 16765
    },
    {
      "epoch": 3.991489614949711,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 16766
    },
    {
      "epoch": 3.991727667678391,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 16767
    },
    {
      "epoch": 3.9919657204070704,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 16768
    },
    {
      "epoch": 3.9922037731357496,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 16769
    },
    {
      "epoch": 3.9924418258644288,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 16770
    },
    {
      "epoch": 3.9926798785931084,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 16771
    },
    {
      "epoch": 3.992917931321788,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 16772
    },
    {
      "epoch": 3.993155984050467,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2967,
      "step": 16773
    },
    {
      "epoch": 3.9933940367791467,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 16774
    },
    {
      "epoch": 3.993632089507826,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 16775
    },
    {
      "epoch": 3.9938701422365055,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 16776
    },
    {
      "epoch": 3.9941081949651847,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 16777
    },
    {
      "epoch": 3.9943462476938643,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 16778
    },
    {
      "epoch": 3.9945843004225434,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 16779
    },
    {
      "epoch": 3.994822353151223,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2158,
      "step": 16780
    },
    {
      "epoch": 3.9950604058799026,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 16781
    },
    {
      "epoch": 3.995298458608582,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 16782
    },
    {
      "epoch": 3.995536511337261,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 16783
    },
    {
      "epoch": 3.9957745640659406,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 16784
    },
    {
      "epoch": 3.99601261679462,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 16785
    },
    {
      "epoch": 3.9962506695232993,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2872,
      "step": 16786
    },
    {
      "epoch": 3.996488722251979,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 16787
    },
    {
      "epoch": 3.996726774980658,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 16788
    },
    {
      "epoch": 3.9969648277093377,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8808,
      "step": 16789
    },
    {
      "epoch": 3.997202880438017,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 16790
    },
    {
      "epoch": 3.9974409331666965,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 16791
    },
    {
      "epoch": 3.9976789858953756,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 16792
    },
    {
      "epoch": 3.9979170386240552,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 16793
    },
    {
      "epoch": 3.998155091352735,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 16794
    },
    {
      "epoch": 3.998393144081414,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 16795
    },
    {
      "epoch": 3.9986311968100936,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 16796
    },
    {
      "epoch": 3.9988692495387728,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.468,
      "step": 16797
    },
    {
      "epoch": 3.9991073022674524,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 16798
    },
    {
      "epoch": 3.9993453549961315,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 16799
    },
    {
      "epoch": 3.999583407724811,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 16800
    },
    {
      "epoch": 3.999583407724811,
      "eval_runtime": 0.5084,
      "eval_samples_per_second": 1.967,
      "eval_steps_per_second": 1.967,
      "step": 16800
    },
    {
      "epoch": 3.9998214604534903,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 16801
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 16802
    },
    {
      "epoch": 4.00023805272868,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 16803
    },
    {
      "epoch": 4.000476105457359,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 16804
    },
    {
      "epoch": 4.000714158186038,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 16805
    },
    {
      "epoch": 4.0009522109147175,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 16806
    },
    {
      "epoch": 4.001190263643397,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 16807
    },
    {
      "epoch": 4.001428316372077,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 16808
    },
    {
      "epoch": 4.0016663691007555,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 16809
    },
    {
      "epoch": 4.001904421829435,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 16810
    },
    {
      "epoch": 4.002142474558115,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 16811
    },
    {
      "epoch": 4.002380527286794,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 16812
    },
    {
      "epoch": 4.002618580015473,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 16813
    },
    {
      "epoch": 4.002856632744153,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 16814
    },
    {
      "epoch": 4.003094685472832,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 16815
    },
    {
      "epoch": 4.003332738201512,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 16816
    },
    {
      "epoch": 4.003570790930191,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 16817
    },
    {
      "epoch": 4.00380884365887,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 16818
    },
    {
      "epoch": 4.00404689638755,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 16819
    },
    {
      "epoch": 4.004284949116229,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 16820
    },
    {
      "epoch": 4.004523001844909,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 16821
    },
    {
      "epoch": 4.004761054573588,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 16822
    },
    {
      "epoch": 4.004999107302267,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 16823
    },
    {
      "epoch": 4.005237160030947,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 16824
    },
    {
      "epoch": 4.0054752127596265,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 16825
    },
    {
      "epoch": 4.005713265488306,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2791,
      "step": 16826
    },
    {
      "epoch": 4.005951318216985,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 16827
    },
    {
      "epoch": 4.006189370945664,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 16828
    },
    {
      "epoch": 4.006427423674344,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3196,
      "step": 16829
    },
    {
      "epoch": 4.006665476403024,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2673,
      "step": 16830
    },
    {
      "epoch": 4.006903529131702,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 16831
    },
    {
      "epoch": 4.007141581860382,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 16832
    },
    {
      "epoch": 4.0073796345890615,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 16833
    },
    {
      "epoch": 4.007617687317741,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 16834
    },
    {
      "epoch": 4.00785574004642,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 16835
    },
    {
      "epoch": 4.0080937927750995,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 16836
    },
    {
      "epoch": 4.008331845503779,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 16837
    },
    {
      "epoch": 4.008569898232459,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 16838
    },
    {
      "epoch": 4.008807950961138,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 16839
    },
    {
      "epoch": 4.009046003689817,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 16840
    },
    {
      "epoch": 4.009284056418497,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 16841
    },
    {
      "epoch": 4.009522109147176,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 16842
    },
    {
      "epoch": 4.009760161875856,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 16843
    },
    {
      "epoch": 4.0099982146045345,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 16844
    },
    {
      "epoch": 4.010236267333214,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 16845
    },
    {
      "epoch": 4.010474320061894,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 16846
    },
    {
      "epoch": 4.010712372790573,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 16847
    },
    {
      "epoch": 4.010950425519252,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 16848
    },
    {
      "epoch": 4.011188478247932,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 16849
    },
    {
      "epoch": 4.011426530976611,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 16850
    },
    {
      "epoch": 4.011664583705291,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 16851
    },
    {
      "epoch": 4.0119026364339705,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 16852
    },
    {
      "epoch": 4.012140689162649,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 16853
    },
    {
      "epoch": 4.012378741891329,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 16854
    },
    {
      "epoch": 4.012616794620008,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 16855
    },
    {
      "epoch": 4.012854847348688,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3568,
      "step": 16856
    },
    {
      "epoch": 4.013092900077367,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2565,
      "step": 16857
    },
    {
      "epoch": 4.013330952806046,
      "grad_norm": 2.625,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 16858
    },
    {
      "epoch": 4.013569005534726,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 16859
    },
    {
      "epoch": 4.013807058263406,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 16860
    },
    {
      "epoch": 4.014045110992085,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 16861
    },
    {
      "epoch": 4.014283163720764,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 16862
    },
    {
      "epoch": 4.0145212164494435,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 16863
    },
    {
      "epoch": 4.014759269178123,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 16864
    },
    {
      "epoch": 4.014997321906803,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 16865
    },
    {
      "epoch": 4.015235374635481,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4,
      "step": 16866
    },
    {
      "epoch": 4.015473427364161,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 16867
    },
    {
      "epoch": 4.015711480092841,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 16868
    },
    {
      "epoch": 4.01594953282152,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 16869
    },
    {
      "epoch": 4.016187585550199,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 16870
    },
    {
      "epoch": 4.0164256382788786,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4414,
      "step": 16871
    },
    {
      "epoch": 4.016663691007558,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 16872
    },
    {
      "epoch": 4.016901743736238,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.76,
      "step": 16873
    },
    {
      "epoch": 4.017139796464917,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 16874
    },
    {
      "epoch": 4.017377849193596,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 16875
    },
    {
      "epoch": 4.017615901922276,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 16876
    },
    {
      "epoch": 4.017853954650955,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 16877
    },
    {
      "epoch": 4.018092007379635,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1457,
      "step": 16878
    },
    {
      "epoch": 4.018330060108314,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 16879
    },
    {
      "epoch": 4.018568112836993,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 16880
    },
    {
      "epoch": 4.018806165565673,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 16881
    },
    {
      "epoch": 4.019044218294352,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 16882
    },
    {
      "epoch": 4.019282271023032,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 16883
    },
    {
      "epoch": 4.019520323751711,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3291,
      "step": 16884
    },
    {
      "epoch": 4.01975837648039,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 16885
    },
    {
      "epoch": 4.01999642920907,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 16886
    },
    {
      "epoch": 4.02023448193775,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5348,
      "step": 16887
    },
    {
      "epoch": 4.020472534666428,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 16888
    },
    {
      "epoch": 4.020710587395108,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2781,
      "step": 16889
    },
    {
      "epoch": 4.0209486401237875,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 16890
    },
    {
      "epoch": 4.021186692852467,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 16891
    },
    {
      "epoch": 4.021424745581146,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 16892
    },
    {
      "epoch": 4.021662798309825,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 16893
    },
    {
      "epoch": 4.021900851038505,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8093,
      "step": 16894
    },
    {
      "epoch": 4.022138903767185,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 16895
    },
    {
      "epoch": 4.022376956495864,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 16896
    },
    {
      "epoch": 4.022615009224543,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 16897
    },
    {
      "epoch": 4.022853061953223,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 16898
    },
    {
      "epoch": 4.023091114681902,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 16899
    },
    {
      "epoch": 4.023329167410582,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 16900
    },
    {
      "epoch": 4.023329167410582,
      "eval_runtime": 0.5321,
      "eval_samples_per_second": 1.879,
      "eval_steps_per_second": 1.879,
      "step": 16900
    },
    {
      "epoch": 4.0235672201392605,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 16901
    },
    {
      "epoch": 4.02380527286794,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 16902
    },
    {
      "epoch": 4.02404332559662,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 16903
    },
    {
      "epoch": 4.024281378325299,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7652,
      "step": 16904
    },
    {
      "epoch": 4.024519431053978,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 16905
    },
    {
      "epoch": 4.024757483782658,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 16906
    },
    {
      "epoch": 4.024995536511337,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 16907
    },
    {
      "epoch": 4.025233589240017,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 16908
    },
    {
      "epoch": 4.0254716419686964,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 16909
    },
    {
      "epoch": 4.025709694697375,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 16910
    },
    {
      "epoch": 4.025947747426055,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 16911
    },
    {
      "epoch": 4.026185800154734,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 16912
    },
    {
      "epoch": 4.026423852883414,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 16913
    },
    {
      "epoch": 4.026661905612093,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 16914
    },
    {
      "epoch": 4.026899958340772,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 16915
    },
    {
      "epoch": 4.027138011069452,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3828,
      "step": 16916
    },
    {
      "epoch": 4.0273760637981315,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 16917
    },
    {
      "epoch": 4.027614116526811,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 16918
    },
    {
      "epoch": 4.02785216925549,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 16919
    },
    {
      "epoch": 4.028090221984169,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 16920
    },
    {
      "epoch": 4.028328274712849,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 16921
    },
    {
      "epoch": 4.028566327441529,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 16922
    },
    {
      "epoch": 4.028804380170207,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 16923
    },
    {
      "epoch": 4.029042432898887,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 16924
    },
    {
      "epoch": 4.029280485627567,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 16925
    },
    {
      "epoch": 4.029518538356246,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 16926
    },
    {
      "epoch": 4.029756591084925,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 16927
    },
    {
      "epoch": 4.0299946438136045,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 16928
    },
    {
      "epoch": 4.030232696542284,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.4721,
      "step": 16929
    },
    {
      "epoch": 4.030470749270964,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 16930
    },
    {
      "epoch": 4.030708801999643,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 16931
    },
    {
      "epoch": 4.030946854728322,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2728,
      "step": 16932
    },
    {
      "epoch": 4.031184907457002,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 16933
    },
    {
      "epoch": 4.031422960185681,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 16934
    },
    {
      "epoch": 4.031661012914361,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 16935
    },
    {
      "epoch": 4.03189906564304,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 16936
    },
    {
      "epoch": 4.032137118371719,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 16937
    },
    {
      "epoch": 4.032375171100399,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 16938
    },
    {
      "epoch": 4.032613223829078,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 16939
    },
    {
      "epoch": 4.032851276557757,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 16940
    },
    {
      "epoch": 4.033089329286437,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 16941
    },
    {
      "epoch": 4.033327382015116,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2643,
      "step": 16942
    },
    {
      "epoch": 4.033565434743796,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 16943
    },
    {
      "epoch": 4.0338034874724755,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 16944
    },
    {
      "epoch": 4.034041540201154,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 16945
    },
    {
      "epoch": 4.034279592929834,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8586,
      "step": 16946
    },
    {
      "epoch": 4.0345176456585135,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 16947
    },
    {
      "epoch": 4.034755698387193,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 16948
    },
    {
      "epoch": 4.034993751115872,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.364,
      "step": 16949
    },
    {
      "epoch": 4.035231803844551,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 16950
    },
    {
      "epoch": 4.035469856573231,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 16951
    },
    {
      "epoch": 4.035707909301911,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 16952
    },
    {
      "epoch": 4.03594596203059,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 16953
    },
    {
      "epoch": 4.036184014759269,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 16954
    },
    {
      "epoch": 4.0364220674879485,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 16955
    },
    {
      "epoch": 4.036660120216628,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 16956
    },
    {
      "epoch": 4.036898172945308,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 16957
    },
    {
      "epoch": 4.0371362256739864,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 16958
    },
    {
      "epoch": 4.037374278402666,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 16959
    },
    {
      "epoch": 4.037612331131346,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 16960
    },
    {
      "epoch": 4.037850383860025,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 16961
    },
    {
      "epoch": 4.038088436588704,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 16962
    },
    {
      "epoch": 4.038326489317384,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 16963
    },
    {
      "epoch": 4.038564542046063,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2614,
      "step": 16964
    },
    {
      "epoch": 4.038802594774743,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 16965
    },
    {
      "epoch": 4.039040647503422,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4211,
      "step": 16966
    },
    {
      "epoch": 4.039278700232101,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.348,
      "step": 16967
    },
    {
      "epoch": 4.039516752960781,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 16968
    },
    {
      "epoch": 4.03975480568946,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 16969
    },
    {
      "epoch": 4.03999285841814,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 16970
    },
    {
      "epoch": 4.040230911146819,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7026,
      "step": 16971
    },
    {
      "epoch": 4.040468963875498,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 16972
    },
    {
      "epoch": 4.040707016604178,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5586,
      "step": 16973
    },
    {
      "epoch": 4.0409450693328575,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 16974
    },
    {
      "epoch": 4.041183122061537,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 16975
    },
    {
      "epoch": 4.041421174790216,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 16976
    },
    {
      "epoch": 4.041659227518895,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 16977
    },
    {
      "epoch": 4.041897280247575,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 16978
    },
    {
      "epoch": 4.042135332976255,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 16979
    },
    {
      "epoch": 4.042373385704933,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 16980
    },
    {
      "epoch": 4.042611438433613,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 16981
    },
    {
      "epoch": 4.0428494911622925,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 16982
    },
    {
      "epoch": 4.043087543890972,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3362,
      "step": 16983
    },
    {
      "epoch": 4.043325596619651,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 16984
    },
    {
      "epoch": 4.0435636493483305,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 16985
    },
    {
      "epoch": 4.04380170207701,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 16986
    },
    {
      "epoch": 4.04403975480569,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 16987
    },
    {
      "epoch": 4.044277807534369,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 16988
    },
    {
      "epoch": 4.044515860263048,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 16989
    },
    {
      "epoch": 4.044753912991728,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0122,
      "step": 16990
    },
    {
      "epoch": 4.044991965720407,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 16991
    },
    {
      "epoch": 4.045230018449087,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 16992
    },
    {
      "epoch": 4.0454680711777655,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 16993
    },
    {
      "epoch": 4.045706123906445,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 16994
    },
    {
      "epoch": 4.045944176635125,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 16995
    },
    {
      "epoch": 4.046182229363804,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 16996
    },
    {
      "epoch": 4.046420282092483,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 16997
    },
    {
      "epoch": 4.046658334821163,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 16998
    },
    {
      "epoch": 4.046896387549842,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 16999
    },
    {
      "epoch": 4.047134440278522,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 17000
    },
    {
      "epoch": 4.047134440278522,
      "eval_runtime": 0.513,
      "eval_samples_per_second": 1.949,
      "eval_steps_per_second": 1.949,
      "step": 17000
    },
    {
      "epoch": 4.0473724930072015,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 17001
    },
    {
      "epoch": 4.04761054573588,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 17002
    },
    {
      "epoch": 4.04784859846456,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 17003
    },
    {
      "epoch": 4.048086651193239,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 17004
    },
    {
      "epoch": 4.048324703921919,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 17005
    },
    {
      "epoch": 4.048562756650598,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 17006
    },
    {
      "epoch": 4.048800809379277,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3763,
      "step": 17007
    },
    {
      "epoch": 4.049038862107957,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 17008
    },
    {
      "epoch": 4.0492769148366365,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 17009
    },
    {
      "epoch": 4.049514967565316,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 17010
    },
    {
      "epoch": 4.049753020293995,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 17011
    },
    {
      "epoch": 4.0499910730226745,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4976,
      "step": 17012
    },
    {
      "epoch": 4.050229125751354,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3015,
      "step": 17013
    },
    {
      "epoch": 4.050467178480034,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 17014
    },
    {
      "epoch": 4.050705231208712,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 17015
    },
    {
      "epoch": 4.050943283937392,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 17016
    },
    {
      "epoch": 4.051181336666072,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 17017
    },
    {
      "epoch": 4.051419389394751,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 17018
    },
    {
      "epoch": 4.05165744212343,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 17019
    },
    {
      "epoch": 4.0518954948521095,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 17020
    },
    {
      "epoch": 4.052133547580789,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 17021
    },
    {
      "epoch": 4.052371600309469,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4497,
      "step": 17022
    },
    {
      "epoch": 4.052609653038148,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 17023
    },
    {
      "epoch": 4.052847705766827,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 17024
    },
    {
      "epoch": 4.053085758495507,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 17025
    },
    {
      "epoch": 4.053323811224186,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 17026
    },
    {
      "epoch": 4.053561863952866,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 17027
    },
    {
      "epoch": 4.053799916681545,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 17028
    },
    {
      "epoch": 4.054037969410224,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 17029
    },
    {
      "epoch": 4.054276022138904,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 17030
    },
    {
      "epoch": 4.054514074867583,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 17031
    },
    {
      "epoch": 4.054752127596262,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 17032
    },
    {
      "epoch": 4.054990180324942,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 17033
    },
    {
      "epoch": 4.055228233053621,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5353,
      "step": 17034
    },
    {
      "epoch": 4.055466285782301,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 17035
    },
    {
      "epoch": 4.0557043385109806,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 17036
    },
    {
      "epoch": 4.055942391239659,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 17037
    },
    {
      "epoch": 4.056180443968339,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 17038
    },
    {
      "epoch": 4.0564184966970185,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2513,
      "step": 17039
    },
    {
      "epoch": 4.056656549425698,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 17040
    },
    {
      "epoch": 4.056894602154377,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 17041
    },
    {
      "epoch": 4.057132654883056,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 17042
    },
    {
      "epoch": 4.057370707611736,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 17043
    },
    {
      "epoch": 4.057608760340416,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1158,
      "step": 17044
    },
    {
      "epoch": 4.057846813069095,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 17045
    },
    {
      "epoch": 4.058084865797774,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 17046
    },
    {
      "epoch": 4.0583229185264535,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 17047
    },
    {
      "epoch": 4.058560971255133,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 17048
    },
    {
      "epoch": 4.058799023983813,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 17049
    },
    {
      "epoch": 4.0590370767124915,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 17050
    },
    {
      "epoch": 4.059275129441171,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.375,
      "step": 17051
    },
    {
      "epoch": 4.059513182169851,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 17052
    },
    {
      "epoch": 4.05975123489853,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7565,
      "step": 17053
    },
    {
      "epoch": 4.059989287627209,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 17054
    },
    {
      "epoch": 4.060227340355889,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 17055
    },
    {
      "epoch": 4.060465393084568,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.5812,
      "step": 17056
    },
    {
      "epoch": 4.060703445813248,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 17057
    },
    {
      "epoch": 4.060941498541927,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 17058
    },
    {
      "epoch": 4.061179551270606,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 17059
    },
    {
      "epoch": 4.061417603999286,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 17060
    },
    {
      "epoch": 4.061655656727965,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 17061
    },
    {
      "epoch": 4.061893709456645,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4419,
      "step": 17062
    },
    {
      "epoch": 4.062131762185324,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 17063
    },
    {
      "epoch": 4.062369814914003,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 17064
    },
    {
      "epoch": 4.062607867642683,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 17065
    },
    {
      "epoch": 4.0628459203713625,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 17066
    },
    {
      "epoch": 4.063083973100042,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 17067
    },
    {
      "epoch": 4.063322025828721,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 17068
    },
    {
      "epoch": 4.0635600785574,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 17069
    },
    {
      "epoch": 4.06379813128608,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 17070
    },
    {
      "epoch": 4.06403618401476,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 17071
    },
    {
      "epoch": 4.064274236743438,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 17072
    },
    {
      "epoch": 4.064512289472118,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 17073
    },
    {
      "epoch": 4.064750342200798,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 17074
    },
    {
      "epoch": 4.064988394929477,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 17075
    },
    {
      "epoch": 4.065226447658156,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 17076
    },
    {
      "epoch": 4.0654645003868355,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 17077
    },
    {
      "epoch": 4.065702553115515,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 17078
    },
    {
      "epoch": 4.065940605844195,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 17079
    },
    {
      "epoch": 4.066178658572874,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 17080
    },
    {
      "epoch": 4.066416711301553,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 17081
    },
    {
      "epoch": 4.066654764030233,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 17082
    },
    {
      "epoch": 4.066892816758912,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 17083
    },
    {
      "epoch": 4.067130869487592,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 17084
    },
    {
      "epoch": 4.0673689222162706,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 17085
    },
    {
      "epoch": 4.06760697494495,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3109,
      "step": 17086
    },
    {
      "epoch": 4.06784502767363,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 17087
    },
    {
      "epoch": 4.068083080402309,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 17088
    },
    {
      "epoch": 4.068321133130988,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 17089
    },
    {
      "epoch": 4.068559185859668,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 17090
    },
    {
      "epoch": 4.068797238588347,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 17091
    },
    {
      "epoch": 4.069035291317027,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 17092
    },
    {
      "epoch": 4.0692733440457065,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4303,
      "step": 17093
    },
    {
      "epoch": 4.069511396774385,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 17094
    },
    {
      "epoch": 4.069749449503065,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 17095
    },
    {
      "epoch": 4.069987502231744,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 17096
    },
    {
      "epoch": 4.070225554960424,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 17097
    },
    {
      "epoch": 4.070463607689103,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 17098
    },
    {
      "epoch": 4.070701660417782,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 17099
    },
    {
      "epoch": 4.070939713146462,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 17100
    },
    {
      "epoch": 4.070939713146462,
      "eval_runtime": 0.5303,
      "eval_samples_per_second": 1.886,
      "eval_steps_per_second": 1.886,
      "step": 17100
    },
    {
      "epoch": 4.071177765875142,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 17101
    },
    {
      "epoch": 4.071415818603821,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 17102
    },
    {
      "epoch": 4.0716538713325,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 17103
    },
    {
      "epoch": 4.0718919240611795,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 17104
    },
    {
      "epoch": 4.072129976789859,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 17105
    },
    {
      "epoch": 4.072368029518539,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 17106
    },
    {
      "epoch": 4.072606082247217,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 17107
    },
    {
      "epoch": 4.072844134975897,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3573,
      "step": 17108
    },
    {
      "epoch": 4.073082187704577,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.5599,
      "step": 17109
    },
    {
      "epoch": 4.073320240433256,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 17110
    },
    {
      "epoch": 4.073558293161935,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 17111
    },
    {
      "epoch": 4.073796345890615,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 17112
    },
    {
      "epoch": 4.074034398619294,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 17113
    },
    {
      "epoch": 4.074272451347974,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 17114
    },
    {
      "epoch": 4.074510504076653,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 17115
    },
    {
      "epoch": 4.074748556805332,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 17116
    },
    {
      "epoch": 4.074986609534012,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 17117
    },
    {
      "epoch": 4.075224662262691,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 17118
    },
    {
      "epoch": 4.075462714991371,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 17119
    },
    {
      "epoch": 4.07570076772005,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 17120
    },
    {
      "epoch": 4.075938820448729,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 17121
    },
    {
      "epoch": 4.076176873177409,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 17122
    },
    {
      "epoch": 4.0764149259060884,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 17123
    },
    {
      "epoch": 4.076652978634767,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 17124
    },
    {
      "epoch": 4.076891031363447,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2037,
      "step": 17125
    },
    {
      "epoch": 4.077129084092126,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 17126
    },
    {
      "epoch": 4.077367136820806,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 17127
    },
    {
      "epoch": 4.077605189549486,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 17128
    },
    {
      "epoch": 4.077843242278164,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 17129
    },
    {
      "epoch": 4.078081295006844,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 17130
    },
    {
      "epoch": 4.0783193477355235,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 17131
    },
    {
      "epoch": 4.078557400464203,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 17132
    },
    {
      "epoch": 4.078795453192882,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 17133
    },
    {
      "epoch": 4.079033505921561,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 17134
    },
    {
      "epoch": 4.079271558650241,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3299,
      "step": 17135
    },
    {
      "epoch": 4.079509611378921,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 17136
    },
    {
      "epoch": 4.0797476641076,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 17137
    },
    {
      "epoch": 4.079985716836279,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3545,
      "step": 17138
    },
    {
      "epoch": 4.080223769564959,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 17139
    },
    {
      "epoch": 4.080461822293638,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 17140
    },
    {
      "epoch": 4.080699875022318,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 17141
    },
    {
      "epoch": 4.0809379277509965,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 17142
    },
    {
      "epoch": 4.081175980479676,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 17143
    },
    {
      "epoch": 4.081414033208356,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 17144
    },
    {
      "epoch": 4.081652085937035,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 17145
    },
    {
      "epoch": 4.081890138665714,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 17146
    },
    {
      "epoch": 4.082128191394394,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 17147
    },
    {
      "epoch": 4.082366244123073,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 17148
    },
    {
      "epoch": 4.082604296851753,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 17149
    },
    {
      "epoch": 4.0828423495804325,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 17150
    },
    {
      "epoch": 4.083080402309111,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 17151
    },
    {
      "epoch": 4.083318455037791,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 17152
    },
    {
      "epoch": 4.08355650776647,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 17153
    },
    {
      "epoch": 4.08379456049515,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 17154
    },
    {
      "epoch": 4.084032613223829,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.283,
      "step": 17155
    },
    {
      "epoch": 4.084270665952508,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 17156
    },
    {
      "epoch": 4.084508718681188,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 17157
    },
    {
      "epoch": 4.0847467714098675,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 17158
    },
    {
      "epoch": 4.084984824138546,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 17159
    },
    {
      "epoch": 4.085222876867226,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4144,
      "step": 17160
    },
    {
      "epoch": 4.0854609295959055,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 17161
    },
    {
      "epoch": 4.085698982324585,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 17162
    },
    {
      "epoch": 4.085937035053265,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.605,
      "step": 17163
    },
    {
      "epoch": 4.086175087781943,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4226,
      "step": 17164
    },
    {
      "epoch": 4.086413140510623,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 17165
    },
    {
      "epoch": 4.086651193239303,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 17166
    },
    {
      "epoch": 4.086889245967982,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 17167
    },
    {
      "epoch": 4.087127298696661,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 17168
    },
    {
      "epoch": 4.0873653514253405,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 17169
    },
    {
      "epoch": 4.08760340415402,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 17170
    },
    {
      "epoch": 4.0878414568827,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 17171
    },
    {
      "epoch": 4.088079509611379,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 17172
    },
    {
      "epoch": 4.088317562340058,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 17173
    },
    {
      "epoch": 4.088555615068738,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8478,
      "step": 17174
    },
    {
      "epoch": 4.088793667797417,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1948,
      "step": 17175
    },
    {
      "epoch": 4.089031720526097,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 17176
    },
    {
      "epoch": 4.089269773254776,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 17177
    },
    {
      "epoch": 4.089507825983455,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 17178
    },
    {
      "epoch": 4.089745878712135,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 17179
    },
    {
      "epoch": 4.089983931440814,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 17180
    },
    {
      "epoch": 4.090221984169493,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 17181
    },
    {
      "epoch": 4.090460036898173,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 17182
    },
    {
      "epoch": 4.090698089626852,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 17183
    },
    {
      "epoch": 4.090936142355532,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 17184
    },
    {
      "epoch": 4.0911741950842115,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 17185
    },
    {
      "epoch": 4.09141224781289,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 17186
    },
    {
      "epoch": 4.09165030054157,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 17187
    },
    {
      "epoch": 4.0918883532702495,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 17188
    },
    {
      "epoch": 4.092126405998929,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 17189
    },
    {
      "epoch": 4.092364458727608,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 17190
    },
    {
      "epoch": 4.092602511456287,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 17191
    },
    {
      "epoch": 4.092840564184967,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 17192
    },
    {
      "epoch": 4.093078616913647,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 17193
    },
    {
      "epoch": 4.093316669642326,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 17194
    },
    {
      "epoch": 4.093554722371005,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 17195
    },
    {
      "epoch": 4.0937927750996845,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 17196
    },
    {
      "epoch": 4.094030827828364,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 17197
    },
    {
      "epoch": 4.094268880557044,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 17198
    },
    {
      "epoch": 4.0945069332857225,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 17199
    },
    {
      "epoch": 4.094744986014402,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 17200
    },
    {
      "epoch": 4.094744986014402,
      "eval_runtime": 0.5241,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 17200
    },
    {
      "epoch": 4.094983038743082,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 17201
    },
    {
      "epoch": 4.095221091471761,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 17202
    },
    {
      "epoch": 4.09545914420044,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 17203
    },
    {
      "epoch": 4.09569719692912,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 17204
    },
    {
      "epoch": 4.095935249657799,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 17205
    },
    {
      "epoch": 4.096173302386479,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 17206
    },
    {
      "epoch": 4.096411355115158,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4075,
      "step": 17207
    },
    {
      "epoch": 4.096649407843837,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 17208
    },
    {
      "epoch": 4.096887460572517,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 17209
    },
    {
      "epoch": 4.097125513301196,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 17210
    },
    {
      "epoch": 4.097363566029876,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 17211
    },
    {
      "epoch": 4.097601618758555,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 17212
    },
    {
      "epoch": 4.097839671487234,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 17213
    },
    {
      "epoch": 4.098077724215914,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 17214
    },
    {
      "epoch": 4.0983157769445935,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 17215
    },
    {
      "epoch": 4.098553829673272,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 17216
    },
    {
      "epoch": 4.098791882401952,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 17217
    },
    {
      "epoch": 4.099029935130631,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 17218
    },
    {
      "epoch": 4.099267987859311,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 17219
    },
    {
      "epoch": 4.099506040587991,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 17220
    },
    {
      "epoch": 4.099744093316669,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 17221
    },
    {
      "epoch": 4.099982146045349,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 17222
    },
    {
      "epoch": 4.1002201987740285,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 17223
    },
    {
      "epoch": 4.100458251502708,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 17224
    },
    {
      "epoch": 4.100696304231387,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 17225
    },
    {
      "epoch": 4.1009343569600665,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 17226
    },
    {
      "epoch": 4.101172409688746,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 17227
    },
    {
      "epoch": 4.101410462417426,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 17228
    },
    {
      "epoch": 4.101648515146105,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 17229
    },
    {
      "epoch": 4.101886567874784,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 17230
    },
    {
      "epoch": 4.102124620603464,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 17231
    },
    {
      "epoch": 4.102362673332143,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 17232
    },
    {
      "epoch": 4.102600726060823,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 17233
    },
    {
      "epoch": 4.1028387787895015,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 17234
    },
    {
      "epoch": 4.103076831518181,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3539,
      "step": 17235
    },
    {
      "epoch": 4.103314884246861,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 17236
    },
    {
      "epoch": 4.10355293697554,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 17237
    },
    {
      "epoch": 4.103790989704219,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 17238
    },
    {
      "epoch": 4.104029042432899,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 17239
    },
    {
      "epoch": 4.104267095161578,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 17240
    },
    {
      "epoch": 4.104505147890258,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 17241
    },
    {
      "epoch": 4.1047432006189375,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 17242
    },
    {
      "epoch": 4.104981253347616,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 17243
    },
    {
      "epoch": 4.105219306076296,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 17244
    },
    {
      "epoch": 4.105457358804975,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 17245
    },
    {
      "epoch": 4.105695411533655,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 17246
    },
    {
      "epoch": 4.105933464262334,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 17247
    },
    {
      "epoch": 4.106171516991013,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 17248
    },
    {
      "epoch": 4.106409569719693,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 17249
    },
    {
      "epoch": 4.106647622448373,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 17250
    },
    {
      "epoch": 4.106885675177052,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 17251
    },
    {
      "epoch": 4.107123727905731,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 17252
    },
    {
      "epoch": 4.1073617806344105,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 17253
    },
    {
      "epoch": 4.10759983336309,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 17254
    },
    {
      "epoch": 4.10783788609177,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 17255
    },
    {
      "epoch": 4.108075938820448,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 17256
    },
    {
      "epoch": 4.108313991549128,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 17257
    },
    {
      "epoch": 4.108552044277808,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 17258
    },
    {
      "epoch": 4.108790097006487,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 17259
    },
    {
      "epoch": 4.109028149735166,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 17260
    },
    {
      "epoch": 4.1092662024638456,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 17261
    },
    {
      "epoch": 4.109504255192525,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 17262
    },
    {
      "epoch": 4.109742307921205,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 17263
    },
    {
      "epoch": 4.109980360649884,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 17264
    },
    {
      "epoch": 4.110218413378563,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 17265
    },
    {
      "epoch": 4.110456466107243,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 17266
    },
    {
      "epoch": 4.110694518835922,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 17267
    },
    {
      "epoch": 4.110932571564602,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 17268
    },
    {
      "epoch": 4.111170624293281,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 17269
    },
    {
      "epoch": 4.11140867702196,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 17270
    },
    {
      "epoch": 4.11164672975064,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 17271
    },
    {
      "epoch": 4.111884782479319,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4319,
      "step": 17272
    },
    {
      "epoch": 4.112122835207998,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 17273
    },
    {
      "epoch": 4.112360887936678,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 17274
    },
    {
      "epoch": 4.112598940665357,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 17275
    },
    {
      "epoch": 4.112836993394037,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 17276
    },
    {
      "epoch": 4.113075046122717,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 17277
    },
    {
      "epoch": 4.113313098851395,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 17278
    },
    {
      "epoch": 4.113551151580075,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 17279
    },
    {
      "epoch": 4.1137892043087545,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 17280
    },
    {
      "epoch": 4.114027257037434,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 17281
    },
    {
      "epoch": 4.114265309766113,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 17282
    },
    {
      "epoch": 4.114503362494792,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 17283
    },
    {
      "epoch": 4.114741415223472,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 17284
    },
    {
      "epoch": 4.114979467952152,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 17285
    },
    {
      "epoch": 4.115217520680831,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 17286
    },
    {
      "epoch": 4.11545557340951,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 17287
    },
    {
      "epoch": 4.11569362613819,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 17288
    },
    {
      "epoch": 4.115931678866869,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 17289
    },
    {
      "epoch": 4.116169731595549,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 17290
    },
    {
      "epoch": 4.1164077843242275,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8631,
      "step": 17291
    },
    {
      "epoch": 4.116645837052907,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 17292
    },
    {
      "epoch": 4.116883889781587,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 17293
    },
    {
      "epoch": 4.117121942510266,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 17294
    },
    {
      "epoch": 4.117359995238945,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 17295
    },
    {
      "epoch": 4.117598047967625,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 17296
    },
    {
      "epoch": 4.117836100696304,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 17297
    },
    {
      "epoch": 4.118074153424984,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 17298
    },
    {
      "epoch": 4.1183122061536634,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 17299
    },
    {
      "epoch": 4.118550258882342,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 17300
    },
    {
      "epoch": 4.118550258882342,
      "eval_runtime": 0.5224,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 17300
    },
    {
      "epoch": 4.118788311611022,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 17301
    },
    {
      "epoch": 4.119026364339701,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3361,
      "step": 17302
    },
    {
      "epoch": 4.119264417068381,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 17303
    },
    {
      "epoch": 4.11950246979706,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 17304
    },
    {
      "epoch": 4.119740522525739,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 17305
    },
    {
      "epoch": 4.119978575254419,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 17306
    },
    {
      "epoch": 4.1202166279830985,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 17307
    },
    {
      "epoch": 4.120454680711777,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 17308
    },
    {
      "epoch": 4.120692733440457,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 17309
    },
    {
      "epoch": 4.120930786169136,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 17310
    },
    {
      "epoch": 4.121168838897816,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2487,
      "step": 17311
    },
    {
      "epoch": 4.121406891626496,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 17312
    },
    {
      "epoch": 4.121644944355174,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 17313
    },
    {
      "epoch": 4.121882997083854,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4292,
      "step": 17314
    },
    {
      "epoch": 4.122121049812534,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3591,
      "step": 17315
    },
    {
      "epoch": 4.122359102541213,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.272,
      "step": 17316
    },
    {
      "epoch": 4.122597155269892,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 17317
    },
    {
      "epoch": 4.1228352079985715,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 17318
    },
    {
      "epoch": 4.123073260727251,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 17319
    },
    {
      "epoch": 4.123311313455931,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 17320
    },
    {
      "epoch": 4.12354936618461,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 17321
    },
    {
      "epoch": 4.123787418913289,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 17322
    },
    {
      "epoch": 4.124025471641969,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 17323
    },
    {
      "epoch": 4.124263524370648,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2946,
      "step": 17324
    },
    {
      "epoch": 4.124501577099328,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 17325
    },
    {
      "epoch": 4.124739629828007,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 17326
    },
    {
      "epoch": 4.124977682556686,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 17327
    },
    {
      "epoch": 4.125215735285366,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 17328
    },
    {
      "epoch": 4.125453788014045,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 17329
    },
    {
      "epoch": 4.125691840742724,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 17330
    },
    {
      "epoch": 4.125929893471404,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 17331
    },
    {
      "epoch": 4.126167946200083,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 17332
    },
    {
      "epoch": 4.126405998928763,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 17333
    },
    {
      "epoch": 4.1266440516574425,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 17334
    },
    {
      "epoch": 4.126882104386121,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 17335
    },
    {
      "epoch": 4.127120157114801,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 17336
    },
    {
      "epoch": 4.1273582098434805,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 17337
    },
    {
      "epoch": 4.12759626257216,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 17338
    },
    {
      "epoch": 4.127834315300839,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 17339
    },
    {
      "epoch": 4.128072368029518,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 17340
    },
    {
      "epoch": 4.128310420758198,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 17341
    },
    {
      "epoch": 4.128548473486878,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 17342
    },
    {
      "epoch": 4.128786526215556,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 17343
    },
    {
      "epoch": 4.129024578944236,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 17344
    },
    {
      "epoch": 4.1292626316729155,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 17345
    },
    {
      "epoch": 4.129500684401595,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 17346
    },
    {
      "epoch": 4.129738737130275,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 17347
    },
    {
      "epoch": 4.1299767898589534,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 17348
    },
    {
      "epoch": 4.130214842587633,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3947,
      "step": 17349
    },
    {
      "epoch": 4.130452895316313,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 17350
    },
    {
      "epoch": 4.130690948044992,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 17351
    },
    {
      "epoch": 4.130929000773671,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 17352
    },
    {
      "epoch": 4.131167053502351,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 17353
    },
    {
      "epoch": 4.13140510623103,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 17354
    },
    {
      "epoch": 4.13164315895971,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 17355
    },
    {
      "epoch": 4.131881211688389,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 17356
    },
    {
      "epoch": 4.132119264417068,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 17357
    },
    {
      "epoch": 4.132357317145748,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 17358
    },
    {
      "epoch": 4.132595369874427,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 17359
    },
    {
      "epoch": 4.132833422603107,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 17360
    },
    {
      "epoch": 4.133071475331786,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 17361
    },
    {
      "epoch": 4.133309528060465,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 17362
    },
    {
      "epoch": 4.133547580789145,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3618,
      "step": 17363
    },
    {
      "epoch": 4.1337856335178245,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 17364
    },
    {
      "epoch": 4.134023686246503,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.835,
      "step": 17365
    },
    {
      "epoch": 4.134261738975183,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 17366
    },
    {
      "epoch": 4.134499791703862,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 17367
    },
    {
      "epoch": 4.134737844432542,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 17368
    },
    {
      "epoch": 4.134975897161222,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 17369
    },
    {
      "epoch": 4.1352139498899,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4635,
      "step": 17370
    },
    {
      "epoch": 4.13545200261858,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 17371
    },
    {
      "epoch": 4.1356900553472595,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 17372
    },
    {
      "epoch": 4.135928108075939,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 17373
    },
    {
      "epoch": 4.136166160804618,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 17374
    },
    {
      "epoch": 4.1364042135332975,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 17375
    },
    {
      "epoch": 4.136642266261977,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 17376
    },
    {
      "epoch": 4.136880318990657,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 17377
    },
    {
      "epoch": 4.137118371719335,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 17378
    },
    {
      "epoch": 4.137356424448015,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 17379
    },
    {
      "epoch": 4.137594477176695,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 17380
    },
    {
      "epoch": 4.137832529905374,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 17381
    },
    {
      "epoch": 4.138070582634054,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 17382
    },
    {
      "epoch": 4.1383086353627325,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 17383
    },
    {
      "epoch": 4.138546688091412,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 17384
    },
    {
      "epoch": 4.138784740820092,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 17385
    },
    {
      "epoch": 4.139022793548771,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 17386
    },
    {
      "epoch": 4.13926084627745,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 17387
    },
    {
      "epoch": 4.13949889900613,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 17388
    },
    {
      "epoch": 4.139736951734809,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 17389
    },
    {
      "epoch": 4.139975004463489,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 17390
    },
    {
      "epoch": 4.1402130571921685,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 17391
    },
    {
      "epoch": 4.140451109920847,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9363,
      "step": 17392
    },
    {
      "epoch": 4.140689162649527,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 17393
    },
    {
      "epoch": 4.140927215378206,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 17394
    },
    {
      "epoch": 4.141165268106886,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 17395
    },
    {
      "epoch": 4.141403320835565,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 17396
    },
    {
      "epoch": 4.141641373564244,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 17397
    },
    {
      "epoch": 4.141879426292924,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 17398
    },
    {
      "epoch": 4.1421174790216035,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 17399
    },
    {
      "epoch": 4.142355531750282,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 17400
    },
    {
      "epoch": 4.142355531750282,
      "eval_runtime": 0.5304,
      "eval_samples_per_second": 1.885,
      "eval_steps_per_second": 1.885,
      "step": 17400
    },
    {
      "epoch": 4.142593584478962,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 17401
    },
    {
      "epoch": 4.1428316372076415,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3742,
      "step": 17402
    },
    {
      "epoch": 4.143069689936321,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 17403
    },
    {
      "epoch": 4.143307742665001,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 17404
    },
    {
      "epoch": 4.143545795393679,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 17405
    },
    {
      "epoch": 4.143783848122359,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 17406
    },
    {
      "epoch": 4.144021900851039,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 17407
    },
    {
      "epoch": 4.144259953579718,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3669,
      "step": 17408
    },
    {
      "epoch": 4.144498006308397,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4224,
      "step": 17409
    },
    {
      "epoch": 4.1447360590370765,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 17410
    },
    {
      "epoch": 4.144974111765756,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 17411
    },
    {
      "epoch": 4.145212164494436,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 17412
    },
    {
      "epoch": 4.1454502172231145,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 17413
    },
    {
      "epoch": 4.145688269951794,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 17414
    },
    {
      "epoch": 4.145926322680474,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3155,
      "step": 17415
    },
    {
      "epoch": 4.146164375409153,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 17416
    },
    {
      "epoch": 4.146402428137833,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 17417
    },
    {
      "epoch": 4.146640480866512,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 17418
    },
    {
      "epoch": 4.146878533595191,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 17419
    },
    {
      "epoch": 4.147116586323871,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 17420
    },
    {
      "epoch": 4.14735463905255,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 17421
    },
    {
      "epoch": 4.147592691781229,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 17422
    },
    {
      "epoch": 4.147830744509909,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 17423
    },
    {
      "epoch": 4.148068797238588,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 17424
    },
    {
      "epoch": 4.148306849967268,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 17425
    },
    {
      "epoch": 4.148544902695948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 17426
    },
    {
      "epoch": 4.148782955424626,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 17427
    },
    {
      "epoch": 4.149021008153306,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 17428
    },
    {
      "epoch": 4.1492590608819855,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 17429
    },
    {
      "epoch": 4.149497113610665,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 17430
    },
    {
      "epoch": 4.149735166339344,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 17431
    },
    {
      "epoch": 4.149973219068023,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.488,
      "step": 17432
    },
    {
      "epoch": 4.150211271796703,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 17433
    },
    {
      "epoch": 4.150449324525383,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 17434
    },
    {
      "epoch": 4.150687377254062,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 17435
    },
    {
      "epoch": 4.150925429982741,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 17436
    },
    {
      "epoch": 4.1511634827114205,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.762,
      "step": 17437
    },
    {
      "epoch": 4.1514015354401,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 17438
    },
    {
      "epoch": 4.15163958816878,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 17439
    },
    {
      "epoch": 4.1518776408974585,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 17440
    },
    {
      "epoch": 4.152115693626138,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 17441
    },
    {
      "epoch": 4.152353746354818,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 17442
    },
    {
      "epoch": 4.152591799083497,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 17443
    },
    {
      "epoch": 4.152829851812176,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 17444
    },
    {
      "epoch": 4.153067904540856,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 17445
    },
    {
      "epoch": 4.153305957269535,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4078,
      "step": 17446
    },
    {
      "epoch": 4.153544009998215,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 17447
    },
    {
      "epoch": 4.153782062726894,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 17448
    },
    {
      "epoch": 4.154020115455573,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 17449
    },
    {
      "epoch": 4.154258168184253,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 17450
    },
    {
      "epoch": 4.154496220912932,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 17451
    },
    {
      "epoch": 4.154734273641612,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3866,
      "step": 17452
    },
    {
      "epoch": 4.154972326370291,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 17453
    },
    {
      "epoch": 4.15521037909897,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 17454
    },
    {
      "epoch": 4.15544843182765,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2485,
      "step": 17455
    },
    {
      "epoch": 4.1556864845563295,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 17456
    },
    {
      "epoch": 4.155924537285008,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 17457
    },
    {
      "epoch": 4.156162590013688,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 17458
    },
    {
      "epoch": 4.156400642742367,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 17459
    },
    {
      "epoch": 4.156638695471047,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 17460
    },
    {
      "epoch": 4.156876748199727,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 17461
    },
    {
      "epoch": 4.157114800928405,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 17462
    },
    {
      "epoch": 4.157352853657085,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 17463
    },
    {
      "epoch": 4.157590906385765,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 17464
    },
    {
      "epoch": 4.157828959114444,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 17465
    },
    {
      "epoch": 4.158067011843123,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 17466
    },
    {
      "epoch": 4.1583050645718025,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.4302,
      "step": 17467
    },
    {
      "epoch": 4.158543117300482,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 17468
    },
    {
      "epoch": 4.158781170029162,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 17469
    },
    {
      "epoch": 4.159019222757841,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 17470
    },
    {
      "epoch": 4.15925727548652,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 17471
    },
    {
      "epoch": 4.1594953282152,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 17472
    },
    {
      "epoch": 4.159733380943879,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 17473
    },
    {
      "epoch": 4.159971433672559,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 17474
    },
    {
      "epoch": 4.1602094864012376,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 17475
    },
    {
      "epoch": 4.160447539129917,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 17476
    },
    {
      "epoch": 4.160685591858597,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 17477
    },
    {
      "epoch": 4.160923644587276,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 17478
    },
    {
      "epoch": 4.161161697315955,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 17479
    },
    {
      "epoch": 4.161399750044635,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1886,
      "step": 17480
    },
    {
      "epoch": 4.161637802773314,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 17481
    },
    {
      "epoch": 4.161875855501994,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 17482
    },
    {
      "epoch": 4.1621139082306735,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4788,
      "step": 17483
    },
    {
      "epoch": 4.162351960959352,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 17484
    },
    {
      "epoch": 4.162590013688032,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 17485
    },
    {
      "epoch": 4.162828066416711,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 17486
    },
    {
      "epoch": 4.163066119145391,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 17487
    },
    {
      "epoch": 4.16330417187407,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2356,
      "step": 17488
    },
    {
      "epoch": 4.163542224602749,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 17489
    },
    {
      "epoch": 4.163780277331429,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 17490
    },
    {
      "epoch": 4.164018330060109,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 17491
    },
    {
      "epoch": 4.164256382788787,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 17492
    },
    {
      "epoch": 4.164494435517467,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 17493
    },
    {
      "epoch": 4.1647324882461465,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3013,
      "step": 17494
    },
    {
      "epoch": 4.164970540974826,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 17495
    },
    {
      "epoch": 4.165208593703506,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 17496
    },
    {
      "epoch": 4.165446646432184,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 17497
    },
    {
      "epoch": 4.165684699160864,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 17498
    },
    {
      "epoch": 4.165922751889544,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 17499
    },
    {
      "epoch": 4.166160804618223,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 17500
    },
    {
      "epoch": 4.166160804618223,
      "eval_runtime": 0.5244,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 17500
    },
    {
      "epoch": 4.166398857346902,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 17501
    },
    {
      "epoch": 4.166636910075582,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 17502
    },
    {
      "epoch": 4.166874962804261,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 17503
    },
    {
      "epoch": 4.167113015532941,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 17504
    },
    {
      "epoch": 4.16735106826162,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5866,
      "step": 17505
    },
    {
      "epoch": 4.167589120990299,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 17506
    },
    {
      "epoch": 4.167827173718979,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 17507
    },
    {
      "epoch": 4.168065226447658,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 17508
    },
    {
      "epoch": 4.168303279176338,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 17509
    },
    {
      "epoch": 4.168541331905017,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 17510
    },
    {
      "epoch": 4.168779384633696,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 17511
    },
    {
      "epoch": 4.169017437362376,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 17512
    },
    {
      "epoch": 4.1692554900910554,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 17513
    },
    {
      "epoch": 4.169493542819734,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 17514
    },
    {
      "epoch": 4.169731595548414,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 17515
    },
    {
      "epoch": 4.169969648277093,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 17516
    },
    {
      "epoch": 4.170207701005773,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 17517
    },
    {
      "epoch": 4.170445753734453,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 17518
    },
    {
      "epoch": 4.170683806463131,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 17519
    },
    {
      "epoch": 4.170921859191811,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 17520
    },
    {
      "epoch": 4.1711599119204905,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 17521
    },
    {
      "epoch": 4.17139796464917,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 17522
    },
    {
      "epoch": 4.171636017377849,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 17523
    },
    {
      "epoch": 4.171874070106528,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 17524
    },
    {
      "epoch": 4.172112122835208,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 17525
    },
    {
      "epoch": 4.172350175563888,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 17526
    },
    {
      "epoch": 4.172588228292566,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 17527
    },
    {
      "epoch": 4.172826281021246,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 17528
    },
    {
      "epoch": 4.173064333749926,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 17529
    },
    {
      "epoch": 4.173302386478605,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4421,
      "step": 17530
    },
    {
      "epoch": 4.173540439207285,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 17531
    },
    {
      "epoch": 4.1737784919359635,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 17532
    },
    {
      "epoch": 4.174016544664643,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 17533
    },
    {
      "epoch": 4.174254597393323,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 17534
    },
    {
      "epoch": 4.174492650122002,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 17535
    },
    {
      "epoch": 4.174730702850681,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 17536
    },
    {
      "epoch": 4.174968755579361,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 17537
    },
    {
      "epoch": 4.17520680830804,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 17538
    },
    {
      "epoch": 4.17544486103672,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 17539
    },
    {
      "epoch": 4.1756829137653995,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 17540
    },
    {
      "epoch": 4.175920966494078,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 17541
    },
    {
      "epoch": 4.176159019222758,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 17542
    },
    {
      "epoch": 4.176397071951437,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 17543
    },
    {
      "epoch": 4.176635124680117,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 17544
    },
    {
      "epoch": 4.176873177408796,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4331,
      "step": 17545
    },
    {
      "epoch": 4.177111230137475,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 17546
    },
    {
      "epoch": 4.177349282866155,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 17547
    },
    {
      "epoch": 4.1775873355948345,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 17548
    },
    {
      "epoch": 4.177825388323513,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 17549
    },
    {
      "epoch": 4.178063441052193,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 17550
    },
    {
      "epoch": 4.1783014937808725,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 17551
    },
    {
      "epoch": 4.178539546509552,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 17552
    },
    {
      "epoch": 4.178777599238232,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 17553
    },
    {
      "epoch": 4.17901565196691,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 17554
    },
    {
      "epoch": 4.17925370469559,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5904,
      "step": 17555
    },
    {
      "epoch": 4.17949175742427,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 17556
    },
    {
      "epoch": 4.179729810152949,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 17557
    },
    {
      "epoch": 4.179967862881628,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 17558
    },
    {
      "epoch": 4.1802059156103075,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 17559
    },
    {
      "epoch": 4.180443968338987,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 17560
    },
    {
      "epoch": 4.180682021067667,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 17561
    },
    {
      "epoch": 4.1809200737963454,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 17562
    },
    {
      "epoch": 4.181158126525025,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 17563
    },
    {
      "epoch": 4.181396179253705,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 17564
    },
    {
      "epoch": 4.181634231982384,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 17565
    },
    {
      "epoch": 4.181872284711064,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 17566
    },
    {
      "epoch": 4.182110337439743,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 17567
    },
    {
      "epoch": 4.182348390168422,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5517,
      "step": 17568
    },
    {
      "epoch": 4.182586442897102,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 17569
    },
    {
      "epoch": 4.182824495625781,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 17570
    },
    {
      "epoch": 4.18306254835446,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 17571
    },
    {
      "epoch": 4.18330060108314,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 17572
    },
    {
      "epoch": 4.183538653811819,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 17573
    },
    {
      "epoch": 4.183776706540499,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 17574
    },
    {
      "epoch": 4.1840147592691785,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 17575
    },
    {
      "epoch": 4.184252811997857,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 17576
    },
    {
      "epoch": 4.184490864726537,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 17577
    },
    {
      "epoch": 4.1847289174552165,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 17578
    },
    {
      "epoch": 4.184966970183896,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 17579
    },
    {
      "epoch": 4.185205022912575,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 17580
    },
    {
      "epoch": 4.185443075641254,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 17581
    },
    {
      "epoch": 4.185681128369934,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 17582
    },
    {
      "epoch": 4.185919181098614,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2487,
      "step": 17583
    },
    {
      "epoch": 4.186157233827292,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 17584
    },
    {
      "epoch": 4.186395286555972,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 17585
    },
    {
      "epoch": 4.1866333392846515,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 17586
    },
    {
      "epoch": 4.186871392013331,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5446,
      "step": 17587
    },
    {
      "epoch": 4.187109444742011,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 17588
    },
    {
      "epoch": 4.1873474974706895,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3587,
      "step": 17589
    },
    {
      "epoch": 4.187585550199369,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 17590
    },
    {
      "epoch": 4.187823602928049,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 17591
    },
    {
      "epoch": 4.188061655656728,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7429,
      "step": 17592
    },
    {
      "epoch": 4.188299708385407,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 17593
    },
    {
      "epoch": 4.188537761114087,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 17594
    },
    {
      "epoch": 4.188775813842766,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.3171,
      "step": 17595
    },
    {
      "epoch": 4.189013866571446,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 17596
    },
    {
      "epoch": 4.1892519193001245,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3663,
      "step": 17597
    },
    {
      "epoch": 4.189489972028804,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 17598
    },
    {
      "epoch": 4.189728024757484,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 17599
    },
    {
      "epoch": 4.189966077486163,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 17600
    },
    {
      "epoch": 4.189966077486163,
      "eval_runtime": 0.522,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 17600
    },
    {
      "epoch": 4.190204130214843,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 17601
    },
    {
      "epoch": 4.190442182943522,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 17602
    },
    {
      "epoch": 4.190680235672201,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 17603
    },
    {
      "epoch": 4.190918288400881,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 17604
    },
    {
      "epoch": 4.1911563411295605,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 17605
    },
    {
      "epoch": 4.191394393858239,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 17606
    },
    {
      "epoch": 4.191632446586919,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 17607
    },
    {
      "epoch": 4.191870499315598,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 17608
    },
    {
      "epoch": 4.192108552044278,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3698,
      "step": 17609
    },
    {
      "epoch": 4.192346604772958,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 17610
    },
    {
      "epoch": 4.192584657501636,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 17611
    },
    {
      "epoch": 4.192822710230316,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 17612
    },
    {
      "epoch": 4.1930607629589955,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 17613
    },
    {
      "epoch": 4.193298815687675,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 17614
    },
    {
      "epoch": 4.193536868416354,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 17615
    },
    {
      "epoch": 4.1937749211450335,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 17616
    },
    {
      "epoch": 4.194012973873713,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 17617
    },
    {
      "epoch": 4.194251026602393,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.288,
      "step": 17618
    },
    {
      "epoch": 4.194489079331072,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 17619
    },
    {
      "epoch": 4.194727132059751,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 17620
    },
    {
      "epoch": 4.194965184788431,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 17621
    },
    {
      "epoch": 4.19520323751711,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 17622
    },
    {
      "epoch": 4.19544129024579,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.4965,
      "step": 17623
    },
    {
      "epoch": 4.1956793429744685,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 17624
    },
    {
      "epoch": 4.195917395703148,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 17625
    },
    {
      "epoch": 4.196155448431828,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 17626
    },
    {
      "epoch": 4.196393501160507,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 17627
    },
    {
      "epoch": 4.196631553889186,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4348,
      "step": 17628
    },
    {
      "epoch": 4.196869606617866,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 17629
    },
    {
      "epoch": 4.197107659346545,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 17630
    },
    {
      "epoch": 4.197345712075225,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.307,
      "step": 17631
    },
    {
      "epoch": 4.1975837648039045,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 17632
    },
    {
      "epoch": 4.197821817532583,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 17633
    },
    {
      "epoch": 4.198059870261263,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 17634
    },
    {
      "epoch": 4.198297922989942,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 17635
    },
    {
      "epoch": 4.198535975718622,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 17636
    },
    {
      "epoch": 4.198774028447301,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 17637
    },
    {
      "epoch": 4.19901208117598,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 17638
    },
    {
      "epoch": 4.19925013390466,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 17639
    },
    {
      "epoch": 4.19948818663334,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 17640
    },
    {
      "epoch": 4.199726239362018,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 17641
    },
    {
      "epoch": 4.199964292090698,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 17642
    },
    {
      "epoch": 4.2002023448193775,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 17643
    },
    {
      "epoch": 4.200440397548057,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2454,
      "step": 17644
    },
    {
      "epoch": 4.200678450276737,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 17645
    },
    {
      "epoch": 4.200916503005415,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 17646
    },
    {
      "epoch": 4.201154555734095,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 17647
    },
    {
      "epoch": 4.201392608462775,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 17648
    },
    {
      "epoch": 4.201630661191454,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2768,
      "step": 17649
    },
    {
      "epoch": 4.201868713920133,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 17650
    },
    {
      "epoch": 4.2021067666488126,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.32,
      "step": 17651
    },
    {
      "epoch": 4.202344819377492,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 17652
    },
    {
      "epoch": 4.202582872106172,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 17653
    },
    {
      "epoch": 4.202820924834851,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 17654
    },
    {
      "epoch": 4.20305897756353,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 17655
    },
    {
      "epoch": 4.20329703029221,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 17656
    },
    {
      "epoch": 4.203535083020889,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3447,
      "step": 17657
    },
    {
      "epoch": 4.203773135749569,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 17658
    },
    {
      "epoch": 4.204011188478248,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 17659
    },
    {
      "epoch": 4.204249241206927,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 17660
    },
    {
      "epoch": 4.204487293935607,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 17661
    },
    {
      "epoch": 4.204725346664286,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 17662
    },
    {
      "epoch": 4.204963399392965,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 17663
    },
    {
      "epoch": 4.205201452121645,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 17664
    },
    {
      "epoch": 4.205439504850324,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 17665
    },
    {
      "epoch": 4.205677557579004,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 17666
    },
    {
      "epoch": 4.205915610307684,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 17667
    },
    {
      "epoch": 4.206153663036362,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 17668
    },
    {
      "epoch": 4.206391715765042,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 17669
    },
    {
      "epoch": 4.2066297684937215,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 17670
    },
    {
      "epoch": 4.206867821222401,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 17671
    },
    {
      "epoch": 4.20710587395108,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 17672
    },
    {
      "epoch": 4.207343926679759,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.6637,
      "step": 17673
    },
    {
      "epoch": 4.207581979408439,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 17674
    },
    {
      "epoch": 4.207820032137119,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 17675
    },
    {
      "epoch": 4.208058084865797,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 17676
    },
    {
      "epoch": 4.208296137594477,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 17677
    },
    {
      "epoch": 4.208534190323157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 17678
    },
    {
      "epoch": 4.208772243051836,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 17679
    },
    {
      "epoch": 4.209010295780516,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 17680
    },
    {
      "epoch": 4.2092483485091945,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 17681
    },
    {
      "epoch": 4.209486401237874,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 17682
    },
    {
      "epoch": 4.209724453966554,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 17683
    },
    {
      "epoch": 4.209962506695233,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 17684
    },
    {
      "epoch": 4.210200559423912,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 17685
    },
    {
      "epoch": 4.210438612152592,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.5491,
      "step": 17686
    },
    {
      "epoch": 4.210676664881271,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 17687
    },
    {
      "epoch": 4.210914717609951,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 17688
    },
    {
      "epoch": 4.2111527703386304,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 17689
    },
    {
      "epoch": 4.211390823067309,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 17690
    },
    {
      "epoch": 4.211628875795989,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 17691
    },
    {
      "epoch": 4.211866928524668,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 17692
    },
    {
      "epoch": 4.212104981253348,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 17693
    },
    {
      "epoch": 4.212343033982027,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 17694
    },
    {
      "epoch": 4.212581086710706,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 17695
    },
    {
      "epoch": 4.212819139439386,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 17696
    },
    {
      "epoch": 4.2130571921680655,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 17697
    },
    {
      "epoch": 4.213295244896744,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 17698
    },
    {
      "epoch": 4.213533297625424,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 17699
    },
    {
      "epoch": 4.213771350354103,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 17700
    },
    {
      "epoch": 4.213771350354103,
      "eval_runtime": 0.5232,
      "eval_samples_per_second": 1.911,
      "eval_steps_per_second": 1.911,
      "step": 17700
    },
    {
      "epoch": 4.214009403082783,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 17701
    },
    {
      "epoch": 4.214247455811463,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 17702
    },
    {
      "epoch": 4.214485508540141,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 17703
    },
    {
      "epoch": 4.214723561268821,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 17704
    },
    {
      "epoch": 4.214961613997501,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 17705
    },
    {
      "epoch": 4.21519966672618,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 17706
    },
    {
      "epoch": 4.215437719454859,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2387,
      "step": 17707
    },
    {
      "epoch": 4.2156757721835385,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 17708
    },
    {
      "epoch": 4.215913824912218,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 17709
    },
    {
      "epoch": 4.216151877640898,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 17710
    },
    {
      "epoch": 4.216389930369576,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 17711
    },
    {
      "epoch": 4.216627983098256,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 17712
    },
    {
      "epoch": 4.216866035826936,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 17713
    },
    {
      "epoch": 4.217104088555615,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 17714
    },
    {
      "epoch": 4.217342141284295,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 17715
    },
    {
      "epoch": 4.217580194012974,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 17716
    },
    {
      "epoch": 4.217818246741653,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 17717
    },
    {
      "epoch": 4.218056299470333,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 17718
    },
    {
      "epoch": 4.218294352199012,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 17719
    },
    {
      "epoch": 4.218532404927691,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 17720
    },
    {
      "epoch": 4.218770457656371,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 17721
    },
    {
      "epoch": 4.21900851038505,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3328,
      "step": 17722
    },
    {
      "epoch": 4.21924656311373,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 17723
    },
    {
      "epoch": 4.2194846158424095,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 17724
    },
    {
      "epoch": 4.219722668571088,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 17725
    },
    {
      "epoch": 4.219960721299768,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3901,
      "step": 17726
    },
    {
      "epoch": 4.2201987740284475,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 17727
    },
    {
      "epoch": 4.220436826757127,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 17728
    },
    {
      "epoch": 4.220674879485806,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2728,
      "step": 17729
    },
    {
      "epoch": 4.220912932214485,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 17730
    },
    {
      "epoch": 4.221150984943165,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 17731
    },
    {
      "epoch": 4.221389037671845,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 17732
    },
    {
      "epoch": 4.221627090400523,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 17733
    },
    {
      "epoch": 4.221865143129203,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 17734
    },
    {
      "epoch": 4.2221031958578825,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 17735
    },
    {
      "epoch": 4.222341248586562,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 17736
    },
    {
      "epoch": 4.222579301315242,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 17737
    },
    {
      "epoch": 4.2228173540439204,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 17738
    },
    {
      "epoch": 4.2230554067726,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 17739
    },
    {
      "epoch": 4.22329345950128,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 17740
    },
    {
      "epoch": 4.223531512229959,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 17741
    },
    {
      "epoch": 4.223769564958638,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 17742
    },
    {
      "epoch": 4.224007617687318,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 17743
    },
    {
      "epoch": 4.224245670415997,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 17744
    },
    {
      "epoch": 4.224483723144677,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 17745
    },
    {
      "epoch": 4.2247217758733555,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.278,
      "step": 17746
    },
    {
      "epoch": 4.224959828602035,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 17747
    },
    {
      "epoch": 4.225197881330715,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 17748
    },
    {
      "epoch": 4.225435934059394,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3977,
      "step": 17749
    },
    {
      "epoch": 4.225673986788074,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 17750
    },
    {
      "epoch": 4.225912039516753,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 17751
    },
    {
      "epoch": 4.226150092245432,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 17752
    },
    {
      "epoch": 4.226388144974112,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 17753
    },
    {
      "epoch": 4.2266261977027915,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 17754
    },
    {
      "epoch": 4.22686425043147,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 17755
    },
    {
      "epoch": 4.22710230316015,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 17756
    },
    {
      "epoch": 4.227340355888829,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 17757
    },
    {
      "epoch": 4.227578408617509,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 17758
    },
    {
      "epoch": 4.227816461346189,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 17759
    },
    {
      "epoch": 4.228054514074867,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 17760
    },
    {
      "epoch": 4.228292566803547,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 17761
    },
    {
      "epoch": 4.2285306195322265,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 17762
    },
    {
      "epoch": 4.228768672260906,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 17763
    },
    {
      "epoch": 4.229006724989585,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 17764
    },
    {
      "epoch": 4.2292447777182645,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.43,
      "step": 17765
    },
    {
      "epoch": 4.229482830446944,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 17766
    },
    {
      "epoch": 4.229720883175624,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 17767
    },
    {
      "epoch": 4.229958935904302,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3842,
      "step": 17768
    },
    {
      "epoch": 4.230196988632982,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 17769
    },
    {
      "epoch": 4.230435041361662,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 17770
    },
    {
      "epoch": 4.230673094090341,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 17771
    },
    {
      "epoch": 4.230911146819021,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 17772
    },
    {
      "epoch": 4.2311491995476995,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 17773
    },
    {
      "epoch": 4.231387252276379,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 17774
    },
    {
      "epoch": 4.231625305005059,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3431,
      "step": 17775
    },
    {
      "epoch": 4.231863357733738,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4935,
      "step": 17776
    },
    {
      "epoch": 4.232101410462417,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 17777
    },
    {
      "epoch": 4.232339463191097,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 17778
    },
    {
      "epoch": 4.232577515919776,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 17779
    },
    {
      "epoch": 4.232815568648456,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 17780
    },
    {
      "epoch": 4.233053621377135,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 17781
    },
    {
      "epoch": 4.233291674105814,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 17782
    },
    {
      "epoch": 4.233529726834494,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3177,
      "step": 17783
    },
    {
      "epoch": 4.233767779563173,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 17784
    },
    {
      "epoch": 4.234005832291853,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3281,
      "step": 17785
    },
    {
      "epoch": 4.234243885020532,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 17786
    },
    {
      "epoch": 4.234481937749211,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 17787
    },
    {
      "epoch": 4.234719990477891,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 17788
    },
    {
      "epoch": 4.2349580432065705,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 17789
    },
    {
      "epoch": 4.235196095935249,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3999,
      "step": 17790
    },
    {
      "epoch": 4.235434148663929,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 17791
    },
    {
      "epoch": 4.2356722013926085,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 17792
    },
    {
      "epoch": 4.235910254121288,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 17793
    },
    {
      "epoch": 4.236148306849968,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 17794
    },
    {
      "epoch": 4.236386359578646,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 17795
    },
    {
      "epoch": 4.236624412307326,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 17796
    },
    {
      "epoch": 4.236862465036006,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 17797
    },
    {
      "epoch": 4.237100517764685,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 17798
    },
    {
      "epoch": 4.237338570493364,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 17799
    },
    {
      "epoch": 4.2375766232220435,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 17800
    },
    {
      "epoch": 4.2375766232220435,
      "eval_runtime": 0.5327,
      "eval_samples_per_second": 1.877,
      "eval_steps_per_second": 1.877,
      "step": 17800
    },
    {
      "epoch": 4.237814675950723,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 17801
    },
    {
      "epoch": 4.238052728679403,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 17802
    },
    {
      "epoch": 4.2382907814080815,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 17803
    },
    {
      "epoch": 4.238528834136761,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 17804
    },
    {
      "epoch": 4.238766886865441,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 17805
    },
    {
      "epoch": 4.23900493959412,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 17806
    },
    {
      "epoch": 4.2392429923228,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2727,
      "step": 17807
    },
    {
      "epoch": 4.239481045051479,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 17808
    },
    {
      "epoch": 4.239719097780158,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 17809
    },
    {
      "epoch": 4.239957150508838,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 17810
    },
    {
      "epoch": 4.240195203237517,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 17811
    },
    {
      "epoch": 4.240433255966196,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 17812
    },
    {
      "epoch": 4.240671308694876,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 17813
    },
    {
      "epoch": 4.240909361423555,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 17814
    },
    {
      "epoch": 4.241147414152235,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 17815
    },
    {
      "epoch": 4.241385466880914,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 17816
    },
    {
      "epoch": 4.241623519609593,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 17817
    },
    {
      "epoch": 4.241861572338273,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 17818
    },
    {
      "epoch": 4.2420996250669525,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 17819
    },
    {
      "epoch": 4.242337677795632,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 17820
    },
    {
      "epoch": 4.242575730524311,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 17821
    },
    {
      "epoch": 4.24281378325299,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 17822
    },
    {
      "epoch": 4.24305183598167,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 17823
    },
    {
      "epoch": 4.24328988871035,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3665,
      "step": 17824
    },
    {
      "epoch": 4.243527941439028,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 17825
    },
    {
      "epoch": 4.243765994167708,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 17826
    },
    {
      "epoch": 4.2440040468963875,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 17827
    },
    {
      "epoch": 4.244242099625067,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 17828
    },
    {
      "epoch": 4.244480152353747,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 17829
    },
    {
      "epoch": 4.2447182050824255,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 17830
    },
    {
      "epoch": 4.244956257811105,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 17831
    },
    {
      "epoch": 4.245194310539785,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 17832
    },
    {
      "epoch": 4.245432363268464,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 17833
    },
    {
      "epoch": 4.245670415997143,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 17834
    },
    {
      "epoch": 4.245908468725823,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 17835
    },
    {
      "epoch": 4.246146521454502,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 17836
    },
    {
      "epoch": 4.246384574183182,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 17837
    },
    {
      "epoch": 4.246622626911861,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3298,
      "step": 17838
    },
    {
      "epoch": 4.24686067964054,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 17839
    },
    {
      "epoch": 4.24709873236922,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 17840
    },
    {
      "epoch": 4.247336785097899,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4156,
      "step": 17841
    },
    {
      "epoch": 4.247574837826579,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 17842
    },
    {
      "epoch": 4.247812890555258,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 17843
    },
    {
      "epoch": 4.248050943283937,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 17844
    },
    {
      "epoch": 4.248288996012617,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 17845
    },
    {
      "epoch": 4.2485270487412965,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 17846
    },
    {
      "epoch": 4.248765101469975,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1537,
      "step": 17847
    },
    {
      "epoch": 4.249003154198655,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 17848
    },
    {
      "epoch": 4.249241206927334,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 17849
    },
    {
      "epoch": 4.249479259656014,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 17850
    },
    {
      "epoch": 4.249717312384694,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 17851
    },
    {
      "epoch": 4.249955365113372,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 17852
    },
    {
      "epoch": 4.250193417842052,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4309,
      "step": 17853
    },
    {
      "epoch": 4.250431470570732,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 17854
    },
    {
      "epoch": 4.250669523299411,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 17855
    },
    {
      "epoch": 4.25090757602809,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 17856
    },
    {
      "epoch": 4.2511456287567695,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 17857
    },
    {
      "epoch": 4.251383681485449,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 17858
    },
    {
      "epoch": 4.251621734214129,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 17859
    },
    {
      "epoch": 4.251859786942807,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 17860
    },
    {
      "epoch": 4.252097839671487,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 17861
    },
    {
      "epoch": 4.252335892400167,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3766,
      "step": 17862
    },
    {
      "epoch": 4.252573945128846,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 17863
    },
    {
      "epoch": 4.252811997857526,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 17864
    },
    {
      "epoch": 4.2530500505862046,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 17865
    },
    {
      "epoch": 4.253288103314884,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 17866
    },
    {
      "epoch": 4.253526156043564,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 17867
    },
    {
      "epoch": 4.253764208772243,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 17868
    },
    {
      "epoch": 4.254002261500922,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 17869
    },
    {
      "epoch": 4.254240314229602,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 17870
    },
    {
      "epoch": 4.254478366958281,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 17871
    },
    {
      "epoch": 4.254716419686961,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 17872
    },
    {
      "epoch": 4.2549544724156405,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 17873
    },
    {
      "epoch": 4.255192525144319,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 17874
    },
    {
      "epoch": 4.255430577872999,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 17875
    },
    {
      "epoch": 4.255668630601678,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 17876
    },
    {
      "epoch": 4.255906683330358,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3973,
      "step": 17877
    },
    {
      "epoch": 4.256144736059037,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 17878
    },
    {
      "epoch": 4.256382788787716,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2561,
      "step": 17879
    },
    {
      "epoch": 4.256620841516396,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3207,
      "step": 17880
    },
    {
      "epoch": 4.256858894245076,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 17881
    },
    {
      "epoch": 4.257096946973754,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 17882
    },
    {
      "epoch": 4.257334999702434,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 17883
    },
    {
      "epoch": 4.2575730524311135,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 17884
    },
    {
      "epoch": 4.257811105159793,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 17885
    },
    {
      "epoch": 4.258049157888472,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 17886
    },
    {
      "epoch": 4.258287210617151,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.6042,
      "step": 17887
    },
    {
      "epoch": 4.258525263345831,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 17888
    },
    {
      "epoch": 4.258763316074511,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3641,
      "step": 17889
    },
    {
      "epoch": 4.25900136880319,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 17890
    },
    {
      "epoch": 4.259239421531869,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 17891
    },
    {
      "epoch": 4.259477474260549,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 17892
    },
    {
      "epoch": 4.259715526989228,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 17893
    },
    {
      "epoch": 4.259953579717908,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 17894
    },
    {
      "epoch": 4.2601916324465865,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 17895
    },
    {
      "epoch": 4.260429685175266,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 17896
    },
    {
      "epoch": 4.260667737903946,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 17897
    },
    {
      "epoch": 4.260905790632625,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 17898
    },
    {
      "epoch": 4.261143843361305,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 17899
    },
    {
      "epoch": 4.261381896089984,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7645,
      "step": 17900
    },
    {
      "epoch": 4.261381896089984,
      "eval_runtime": 0.5309,
      "eval_samples_per_second": 1.884,
      "eval_steps_per_second": 1.884,
      "step": 17900
    },
    {
      "epoch": 4.261619948818663,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 17901
    },
    {
      "epoch": 4.261858001547343,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 17902
    },
    {
      "epoch": 4.2620960542760224,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 17903
    },
    {
      "epoch": 4.262334107004701,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 17904
    },
    {
      "epoch": 4.262572159733381,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 17905
    },
    {
      "epoch": 4.26281021246206,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 17906
    },
    {
      "epoch": 4.26304826519074,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 17907
    },
    {
      "epoch": 4.26328631791942,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 17908
    },
    {
      "epoch": 4.263524370648098,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 17909
    },
    {
      "epoch": 4.263762423376778,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 17910
    },
    {
      "epoch": 4.2640004761054575,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 17911
    },
    {
      "epoch": 4.264238528834137,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 17912
    },
    {
      "epoch": 4.264476581562816,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 17913
    },
    {
      "epoch": 4.264714634291495,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 17914
    },
    {
      "epoch": 4.264952687020175,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 17915
    },
    {
      "epoch": 4.265190739748855,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3932,
      "step": 17916
    },
    {
      "epoch": 4.265428792477533,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 17917
    },
    {
      "epoch": 4.265666845206213,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 17918
    },
    {
      "epoch": 4.265904897934893,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 17919
    },
    {
      "epoch": 4.266142950663572,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 17920
    },
    {
      "epoch": 4.266381003392252,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 17921
    },
    {
      "epoch": 4.2666190561209305,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 17922
    },
    {
      "epoch": 4.26685710884961,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 17923
    },
    {
      "epoch": 4.26709516157829,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3302,
      "step": 17924
    },
    {
      "epoch": 4.267333214306969,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 17925
    },
    {
      "epoch": 4.267571267035648,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9533,
      "step": 17926
    },
    {
      "epoch": 4.267809319764328,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9326,
      "step": 17927
    },
    {
      "epoch": 4.268047372493007,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 17928
    },
    {
      "epoch": 4.268285425221687,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 17929
    },
    {
      "epoch": 4.268523477950366,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 17930
    },
    {
      "epoch": 4.268761530679045,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 17931
    },
    {
      "epoch": 4.268999583407725,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 17932
    },
    {
      "epoch": 4.269237636136404,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 17933
    },
    {
      "epoch": 4.269475688865084,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 17934
    },
    {
      "epoch": 4.269713741593763,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 17935
    },
    {
      "epoch": 4.269951794322442,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 17936
    },
    {
      "epoch": 4.270189847051122,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5795,
      "step": 17937
    },
    {
      "epoch": 4.2704278997798015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 17938
    },
    {
      "epoch": 4.27066595250848,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 17939
    },
    {
      "epoch": 4.27090400523716,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 17940
    },
    {
      "epoch": 4.2711420579658395,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 17941
    },
    {
      "epoch": 4.271380110694519,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 17942
    },
    {
      "epoch": 4.271618163423199,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 17943
    },
    {
      "epoch": 4.271856216151877,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 17944
    },
    {
      "epoch": 4.272094268880557,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 17945
    },
    {
      "epoch": 4.272332321609237,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 17946
    },
    {
      "epoch": 4.272570374337916,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 17947
    },
    {
      "epoch": 4.272808427066595,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 17948
    },
    {
      "epoch": 4.2730464797952745,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 17949
    },
    {
      "epoch": 4.273284532523954,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 17950
    },
    {
      "epoch": 4.273522585252634,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 17951
    },
    {
      "epoch": 4.2737606379813124,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 17952
    },
    {
      "epoch": 4.273998690709992,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 17953
    },
    {
      "epoch": 4.274236743438672,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 17954
    },
    {
      "epoch": 4.274474796167351,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 17955
    },
    {
      "epoch": 4.274712848896031,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 17956
    },
    {
      "epoch": 4.27495090162471,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 17957
    },
    {
      "epoch": 4.275188954353389,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 17958
    },
    {
      "epoch": 4.275427007082069,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 17959
    },
    {
      "epoch": 4.275665059810748,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 17960
    },
    {
      "epoch": 4.275903112539427,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 17961
    },
    {
      "epoch": 4.276141165268107,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 17962
    },
    {
      "epoch": 4.276379217996786,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 17963
    },
    {
      "epoch": 4.276617270725466,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 17964
    },
    {
      "epoch": 4.276855323454145,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 17965
    },
    {
      "epoch": 4.277093376182824,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 17966
    },
    {
      "epoch": 4.277331428911504,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 17967
    },
    {
      "epoch": 4.2775694816401835,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2331,
      "step": 17968
    },
    {
      "epoch": 4.277807534368863,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 17969
    },
    {
      "epoch": 4.278045587097542,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 17970
    },
    {
      "epoch": 4.278283639826221,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 17971
    },
    {
      "epoch": 4.278521692554901,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2901,
      "step": 17972
    },
    {
      "epoch": 4.278759745283581,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 17973
    },
    {
      "epoch": 4.278997798012259,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 17974
    },
    {
      "epoch": 4.279235850740939,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 17975
    },
    {
      "epoch": 4.2794739034696185,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 17976
    },
    {
      "epoch": 4.279711956198298,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 17977
    },
    {
      "epoch": 4.279950008926978,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 17978
    },
    {
      "epoch": 4.2801880616556565,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 17979
    },
    {
      "epoch": 4.280426114384336,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 17980
    },
    {
      "epoch": 4.280664167113016,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 17981
    },
    {
      "epoch": 4.280902219841695,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 17982
    },
    {
      "epoch": 4.281140272570374,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 17983
    },
    {
      "epoch": 4.281378325299054,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 17984
    },
    {
      "epoch": 4.281616378027733,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.5846,
      "step": 17985
    },
    {
      "epoch": 4.281854430756413,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 17986
    },
    {
      "epoch": 4.282092483485092,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 17987
    },
    {
      "epoch": 4.282330536213771,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 17988
    },
    {
      "epoch": 4.282568588942451,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 17989
    },
    {
      "epoch": 4.28280664167113,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 17990
    },
    {
      "epoch": 4.28304469439981,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 17991
    },
    {
      "epoch": 4.283282747128489,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 17992
    },
    {
      "epoch": 4.283520799857168,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 17993
    },
    {
      "epoch": 4.283758852585848,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3412,
      "step": 17994
    },
    {
      "epoch": 4.2839969053145275,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 17995
    },
    {
      "epoch": 4.284234958043206,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 17996
    },
    {
      "epoch": 4.284473010771886,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 17997
    },
    {
      "epoch": 4.284711063500565,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 17998
    },
    {
      "epoch": 4.284949116229245,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 17999
    },
    {
      "epoch": 4.285187168957924,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 18000
    },
    {
      "epoch": 4.285187168957924,
      "eval_runtime": 0.523,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 18000
    },
    {
      "epoch": 4.285425221686603,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 18001
    },
    {
      "epoch": 4.285663274415283,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 18002
    },
    {
      "epoch": 4.2859013271439625,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 18003
    },
    {
      "epoch": 4.286139379872642,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 18004
    },
    {
      "epoch": 4.286377432601321,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 18005
    },
    {
      "epoch": 4.2866154853300005,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 18006
    },
    {
      "epoch": 4.28685353805868,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4013,
      "step": 18007
    },
    {
      "epoch": 4.28709159078736,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 18008
    },
    {
      "epoch": 4.287329643516038,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 18009
    },
    {
      "epoch": 4.287567696244718,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 18010
    },
    {
      "epoch": 4.287805748973398,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 18011
    },
    {
      "epoch": 4.288043801702077,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 18012
    },
    {
      "epoch": 4.288281854430757,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 18013
    },
    {
      "epoch": 4.2885199071594355,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3476,
      "step": 18014
    },
    {
      "epoch": 4.288757959888115,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5491,
      "step": 18015
    },
    {
      "epoch": 4.288996012616795,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5677,
      "step": 18016
    },
    {
      "epoch": 4.289234065345474,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 18017
    },
    {
      "epoch": 4.289472118074153,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 18018
    },
    {
      "epoch": 4.289710170802833,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 18019
    },
    {
      "epoch": 4.289948223531512,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 18020
    },
    {
      "epoch": 4.290186276260192,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 18021
    },
    {
      "epoch": 4.2904243289888715,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 18022
    },
    {
      "epoch": 4.29066238171755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 18023
    },
    {
      "epoch": 4.29090043444623,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3462,
      "step": 18024
    },
    {
      "epoch": 4.291138487174909,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 18025
    },
    {
      "epoch": 4.291376539903589,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 18026
    },
    {
      "epoch": 4.291614592632268,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 18027
    },
    {
      "epoch": 4.291852645360947,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 18028
    },
    {
      "epoch": 4.292090698089627,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 18029
    },
    {
      "epoch": 4.292328750818307,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.3803,
      "step": 18030
    },
    {
      "epoch": 4.292566803546985,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 18031
    },
    {
      "epoch": 4.292804856275665,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 18032
    },
    {
      "epoch": 4.2930429090043445,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 18033
    },
    {
      "epoch": 4.293280961733024,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 18034
    },
    {
      "epoch": 4.293519014461703,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2707,
      "step": 18035
    },
    {
      "epoch": 4.293757067190382,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 18036
    },
    {
      "epoch": 4.293995119919062,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 18037
    },
    {
      "epoch": 4.294233172647742,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 18038
    },
    {
      "epoch": 4.294471225376421,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 18039
    },
    {
      "epoch": 4.2947092781051,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.1205,
      "step": 18040
    },
    {
      "epoch": 4.2949473308337796,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.6338,
      "step": 18041
    },
    {
      "epoch": 4.295185383562459,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3621,
      "step": 18042
    },
    {
      "epoch": 4.295423436291139,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3183,
      "step": 18043
    },
    {
      "epoch": 4.2956614890198175,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 18044
    },
    {
      "epoch": 4.295899541748497,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 18045
    },
    {
      "epoch": 4.296137594477177,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 18046
    },
    {
      "epoch": 4.296375647205856,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 18047
    },
    {
      "epoch": 4.296613699934536,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 18048
    },
    {
      "epoch": 4.296851752663215,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2654,
      "step": 18049
    },
    {
      "epoch": 4.297089805391894,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 18050
    },
    {
      "epoch": 4.297327858120574,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 18051
    },
    {
      "epoch": 4.297565910849253,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 18052
    },
    {
      "epoch": 4.297803963577932,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 18053
    },
    {
      "epoch": 4.298042016306612,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 18054
    },
    {
      "epoch": 4.298280069035291,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 18055
    },
    {
      "epoch": 4.298518121763971,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 18056
    },
    {
      "epoch": 4.298756174492651,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 18057
    },
    {
      "epoch": 4.298994227221329,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 18058
    },
    {
      "epoch": 4.299232279950009,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 18059
    },
    {
      "epoch": 4.2994703326786885,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 18060
    },
    {
      "epoch": 4.299708385407368,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 18061
    },
    {
      "epoch": 4.299946438136047,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 18062
    },
    {
      "epoch": 4.300184490864726,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 18063
    },
    {
      "epoch": 4.300422543593406,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 18064
    },
    {
      "epoch": 4.300660596322086,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 18065
    },
    {
      "epoch": 4.300898649050764,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3304,
      "step": 18066
    },
    {
      "epoch": 4.301136701779444,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 18067
    },
    {
      "epoch": 4.301374754508124,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 18068
    },
    {
      "epoch": 4.301612807236803,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 18069
    },
    {
      "epoch": 4.301850859965482,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 18070
    },
    {
      "epoch": 4.3020889126941615,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 18071
    },
    {
      "epoch": 4.302326965422841,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 18072
    },
    {
      "epoch": 4.302565018151521,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 18073
    },
    {
      "epoch": 4.3028030708802,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 18074
    },
    {
      "epoch": 4.303041123608879,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 18075
    },
    {
      "epoch": 4.303279176337559,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4242,
      "step": 18076
    },
    {
      "epoch": 4.303517229066238,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 18077
    },
    {
      "epoch": 4.303755281794918,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 18078
    },
    {
      "epoch": 4.303993334523597,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 18079
    },
    {
      "epoch": 4.304231387252276,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 18080
    },
    {
      "epoch": 4.304469439980956,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 18081
    },
    {
      "epoch": 4.304707492709635,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 18082
    },
    {
      "epoch": 4.304945545438315,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 18083
    },
    {
      "epoch": 4.305183598166994,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 18084
    },
    {
      "epoch": 4.305421650895673,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 18085
    },
    {
      "epoch": 4.305659703624353,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 18086
    },
    {
      "epoch": 4.3058977563530325,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 18087
    },
    {
      "epoch": 4.306135809081711,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 18088
    },
    {
      "epoch": 4.306373861810391,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 18089
    },
    {
      "epoch": 4.30661191453907,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 18090
    },
    {
      "epoch": 4.30684996726775,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 18091
    },
    {
      "epoch": 4.30708801999643,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 18092
    },
    {
      "epoch": 4.307326072725108,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 18093
    },
    {
      "epoch": 4.307564125453788,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 18094
    },
    {
      "epoch": 4.307802178182468,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 18095
    },
    {
      "epoch": 4.308040230911147,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 18096
    },
    {
      "epoch": 4.308278283639826,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 18097
    },
    {
      "epoch": 4.3085163363685055,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 18098
    },
    {
      "epoch": 4.308754389097185,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 18099
    },
    {
      "epoch": 4.308992441825865,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 18100
    },
    {
      "epoch": 4.308992441825865,
      "eval_runtime": 0.5563,
      "eval_samples_per_second": 1.797,
      "eval_steps_per_second": 1.797,
      "step": 18100
    },
    {
      "epoch": 4.309230494554543,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 18101
    },
    {
      "epoch": 4.309468547283223,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 18102
    },
    {
      "epoch": 4.309706600011903,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 18103
    },
    {
      "epoch": 4.309944652740582,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 18104
    },
    {
      "epoch": 4.310182705469262,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.289,
      "step": 18105
    },
    {
      "epoch": 4.310420758197941,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 18106
    },
    {
      "epoch": 4.31065881092662,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 18107
    },
    {
      "epoch": 4.3108968636553,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 18108
    },
    {
      "epoch": 4.311134916383979,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2823,
      "step": 18109
    },
    {
      "epoch": 4.311372969112658,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.4329,
      "step": 18110
    },
    {
      "epoch": 4.311611021841338,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 18111
    },
    {
      "epoch": 4.311849074570017,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 18112
    },
    {
      "epoch": 4.312087127298697,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 18113
    },
    {
      "epoch": 4.312325180027376,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 18114
    },
    {
      "epoch": 4.312563232756055,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 18115
    },
    {
      "epoch": 4.312801285484735,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 18116
    },
    {
      "epoch": 4.3130393382134145,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 18117
    },
    {
      "epoch": 4.313277390942094,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 18118
    },
    {
      "epoch": 4.313515443670773,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 18119
    },
    {
      "epoch": 4.313753496399452,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 18120
    },
    {
      "epoch": 4.313991549128132,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 18121
    },
    {
      "epoch": 4.314229601856812,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 18122
    },
    {
      "epoch": 4.31446765458549,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4492,
      "step": 18123
    },
    {
      "epoch": 4.31470570731417,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 18124
    },
    {
      "epoch": 4.3149437600428495,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 18125
    },
    {
      "epoch": 4.315181812771529,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 18126
    },
    {
      "epoch": 4.315419865500209,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7774,
      "step": 18127
    },
    {
      "epoch": 4.3156579182288874,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 18128
    },
    {
      "epoch": 4.315895970957567,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3786,
      "step": 18129
    },
    {
      "epoch": 4.316134023686247,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 18130
    },
    {
      "epoch": 4.316372076414926,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2549,
      "step": 18131
    },
    {
      "epoch": 4.316610129143605,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 18132
    },
    {
      "epoch": 4.316848181872285,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4223,
      "step": 18133
    },
    {
      "epoch": 4.317086234600964,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 18134
    },
    {
      "epoch": 4.317324287329644,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 18135
    },
    {
      "epoch": 4.3175623400583225,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 18136
    },
    {
      "epoch": 4.317800392787002,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 18137
    },
    {
      "epoch": 4.318038445515682,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 18138
    },
    {
      "epoch": 4.318276498244361,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.5305,
      "step": 18139
    },
    {
      "epoch": 4.318514550973041,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 18140
    },
    {
      "epoch": 4.31875260370172,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 18141
    },
    {
      "epoch": 4.318990656430399,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 18142
    },
    {
      "epoch": 4.319228709159079,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3447,
      "step": 18143
    },
    {
      "epoch": 4.3194667618877585,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 18144
    },
    {
      "epoch": 4.319704814616437,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 18145
    },
    {
      "epoch": 4.319942867345117,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 18146
    },
    {
      "epoch": 4.320180920073796,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 18147
    },
    {
      "epoch": 4.320418972802476,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 18148
    },
    {
      "epoch": 4.320657025531155,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 18149
    },
    {
      "epoch": 4.320895078259834,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 18150
    },
    {
      "epoch": 4.321133130988514,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 18151
    },
    {
      "epoch": 4.3213711837171935,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 18152
    },
    {
      "epoch": 4.321609236445873,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 18153
    },
    {
      "epoch": 4.321847289174552,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 18154
    },
    {
      "epoch": 4.3220853419032315,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3753,
      "step": 18155
    },
    {
      "epoch": 4.322323394631911,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 18156
    },
    {
      "epoch": 4.322561447360591,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.366,
      "step": 18157
    },
    {
      "epoch": 4.322799500089269,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 18158
    },
    {
      "epoch": 4.323037552817949,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 18159
    },
    {
      "epoch": 4.323275605546629,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 18160
    },
    {
      "epoch": 4.323513658275308,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 18161
    },
    {
      "epoch": 4.323751711003988,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 18162
    },
    {
      "epoch": 4.3239897637326665,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 18163
    },
    {
      "epoch": 4.324227816461346,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7272,
      "step": 18164
    },
    {
      "epoch": 4.324465869190026,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 18165
    },
    {
      "epoch": 4.324703921918705,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 18166
    },
    {
      "epoch": 4.324941974647384,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 18167
    },
    {
      "epoch": 4.325180027376064,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 18168
    },
    {
      "epoch": 4.325418080104743,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 18169
    },
    {
      "epoch": 4.325656132833423,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 18170
    },
    {
      "epoch": 4.325894185562102,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 18171
    },
    {
      "epoch": 4.326132238290781,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 18172
    },
    {
      "epoch": 4.326370291019461,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 18173
    },
    {
      "epoch": 4.32660834374814,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 18174
    },
    {
      "epoch": 4.32684639647682,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 18175
    },
    {
      "epoch": 4.327084449205499,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.445,
      "step": 18176
    },
    {
      "epoch": 4.327322501934178,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 18177
    },
    {
      "epoch": 4.327560554662858,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 18178
    },
    {
      "epoch": 4.3277986073915375,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 18179
    },
    {
      "epoch": 4.328036660120216,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 18180
    },
    {
      "epoch": 4.328274712848896,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 18181
    },
    {
      "epoch": 4.3285127655775755,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 18182
    },
    {
      "epoch": 4.328750818306255,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 18183
    },
    {
      "epoch": 4.328988871034934,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3393,
      "step": 18184
    },
    {
      "epoch": 4.329226923763613,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3811,
      "step": 18185
    },
    {
      "epoch": 4.329464976492293,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 18186
    },
    {
      "epoch": 4.329703029220973,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 18187
    },
    {
      "epoch": 4.329941081949652,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 18188
    },
    {
      "epoch": 4.330179134678331,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 18189
    },
    {
      "epoch": 4.3304171874070105,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 18190
    },
    {
      "epoch": 4.33065524013569,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 18191
    },
    {
      "epoch": 4.33089329286437,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3917,
      "step": 18192
    },
    {
      "epoch": 4.3311313455930485,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 18193
    },
    {
      "epoch": 4.331369398321728,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 18194
    },
    {
      "epoch": 4.331607451050408,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 18195
    },
    {
      "epoch": 4.331845503779087,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3969,
      "step": 18196
    },
    {
      "epoch": 4.332083556507767,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 18197
    },
    {
      "epoch": 4.332321609236446,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 18198
    },
    {
      "epoch": 4.332559661965125,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 18199
    },
    {
      "epoch": 4.332797714693805,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 18200
    },
    {
      "epoch": 4.332797714693805,
      "eval_runtime": 0.5164,
      "eval_samples_per_second": 1.937,
      "eval_steps_per_second": 1.937,
      "step": 18200
    },
    {
      "epoch": 4.333035767422484,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 18201
    },
    {
      "epoch": 4.333273820151163,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 18202
    },
    {
      "epoch": 4.333511872879843,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3708,
      "step": 18203
    },
    {
      "epoch": 4.333749925608522,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 18204
    },
    {
      "epoch": 4.333987978337202,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 18205
    },
    {
      "epoch": 4.334226031065882,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 18206
    },
    {
      "epoch": 4.33446408379456,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4963,
      "step": 18207
    },
    {
      "epoch": 4.33470213652324,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 18208
    },
    {
      "epoch": 4.3349401892519195,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 18209
    },
    {
      "epoch": 4.335178241980599,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 18210
    },
    {
      "epoch": 4.335416294709278,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 18211
    },
    {
      "epoch": 4.335654347437957,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.7289,
      "step": 18212
    },
    {
      "epoch": 4.335892400166637,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.5163,
      "step": 18213
    },
    {
      "epoch": 4.336130452895317,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 18214
    },
    {
      "epoch": 4.336368505623995,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 18215
    },
    {
      "epoch": 4.336606558352675,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 18216
    },
    {
      "epoch": 4.3368446110813546,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4937,
      "step": 18217
    },
    {
      "epoch": 4.337082663810034,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 18218
    },
    {
      "epoch": 4.337320716538713,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 18219
    },
    {
      "epoch": 4.3375587692673925,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 18220
    },
    {
      "epoch": 4.337796821996072,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 18221
    },
    {
      "epoch": 4.338034874724752,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 18222
    },
    {
      "epoch": 4.338272927453431,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 18223
    },
    {
      "epoch": 4.33851098018211,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 18224
    },
    {
      "epoch": 4.33874903291079,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 18225
    },
    {
      "epoch": 4.338987085639469,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3753,
      "step": 18226
    },
    {
      "epoch": 4.339225138368149,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 18227
    },
    {
      "epoch": 4.3394631910968275,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 18228
    },
    {
      "epoch": 4.339701243825507,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 18229
    },
    {
      "epoch": 4.339939296554187,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 18230
    },
    {
      "epoch": 4.340177349282866,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 18231
    },
    {
      "epoch": 4.340415402011546,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 18232
    },
    {
      "epoch": 4.340653454740225,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2744,
      "step": 18233
    },
    {
      "epoch": 4.340891507468904,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3082,
      "step": 18234
    },
    {
      "epoch": 4.341129560197584,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 18235
    },
    {
      "epoch": 4.3413676129262635,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 18236
    },
    {
      "epoch": 4.341605665654942,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 18237
    },
    {
      "epoch": 4.341843718383622,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 18238
    },
    {
      "epoch": 4.342081771112301,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 18239
    },
    {
      "epoch": 4.342319823840981,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 18240
    },
    {
      "epoch": 4.342557876569661,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 18241
    },
    {
      "epoch": 4.342795929298339,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 18242
    },
    {
      "epoch": 4.343033982027019,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 18243
    },
    {
      "epoch": 4.343272034755699,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1574,
      "step": 18244
    },
    {
      "epoch": 4.343510087484378,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 18245
    },
    {
      "epoch": 4.343748140213057,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 18246
    },
    {
      "epoch": 4.3439861929417365,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 18247
    },
    {
      "epoch": 4.344224245670416,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 18248
    },
    {
      "epoch": 4.344462298399096,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 18249
    },
    {
      "epoch": 4.344700351127774,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 18250
    },
    {
      "epoch": 4.344938403856454,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 18251
    },
    {
      "epoch": 4.345176456585134,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 18252
    },
    {
      "epoch": 4.345414509313813,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 18253
    },
    {
      "epoch": 4.345652562042492,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 18254
    },
    {
      "epoch": 4.345890614771172,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 18255
    },
    {
      "epoch": 4.346128667499851,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 18256
    },
    {
      "epoch": 4.346366720228531,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 18257
    },
    {
      "epoch": 4.34660477295721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 18258
    },
    {
      "epoch": 4.346842825685889,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 18259
    },
    {
      "epoch": 4.347080878414569,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 18260
    },
    {
      "epoch": 4.347318931143248,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7504,
      "step": 18261
    },
    {
      "epoch": 4.347556983871928,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 18262
    },
    {
      "epoch": 4.347795036600607,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 18263
    },
    {
      "epoch": 4.348033089329286,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 18264
    },
    {
      "epoch": 4.348271142057966,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 18265
    },
    {
      "epoch": 4.348509194786645,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 18266
    },
    {
      "epoch": 4.348747247515325,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 18267
    },
    {
      "epoch": 4.348985300244004,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.5716,
      "step": 18268
    },
    {
      "epoch": 4.349223352972683,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 18269
    },
    {
      "epoch": 4.349461405701363,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 18270
    },
    {
      "epoch": 4.349699458430043,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 18271
    },
    {
      "epoch": 4.349937511158721,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 18272
    },
    {
      "epoch": 4.350175563887401,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 18273
    },
    {
      "epoch": 4.3504136166160805,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 18274
    },
    {
      "epoch": 4.35065166934476,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 18275
    },
    {
      "epoch": 4.35088972207344,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 18276
    },
    {
      "epoch": 4.351127774802118,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 18277
    },
    {
      "epoch": 4.351365827530798,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 18278
    },
    {
      "epoch": 4.351603880259478,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 18279
    },
    {
      "epoch": 4.351841932988157,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 18280
    },
    {
      "epoch": 4.352079985716836,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 18281
    },
    {
      "epoch": 4.352318038445516,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 18282
    },
    {
      "epoch": 4.352556091174195,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7894,
      "step": 18283
    },
    {
      "epoch": 4.352794143902875,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 18284
    },
    {
      "epoch": 4.3530321966315535,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 18285
    },
    {
      "epoch": 4.353270249360233,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 18286
    },
    {
      "epoch": 4.353508302088913,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 18287
    },
    {
      "epoch": 4.353746354817592,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 18288
    },
    {
      "epoch": 4.353984407546271,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 18289
    },
    {
      "epoch": 4.354222460274951,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 18290
    },
    {
      "epoch": 4.35446051300363,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 18291
    },
    {
      "epoch": 4.35469856573231,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 18292
    },
    {
      "epoch": 4.3549366184609895,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 18293
    },
    {
      "epoch": 4.355174671189668,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 18294
    },
    {
      "epoch": 4.355412723918348,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 18295
    },
    {
      "epoch": 4.355650776647027,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 18296
    },
    {
      "epoch": 4.355888829375707,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.876,
      "step": 18297
    },
    {
      "epoch": 4.356126882104386,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3991,
      "step": 18298
    },
    {
      "epoch": 4.356364934833065,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 18299
    },
    {
      "epoch": 4.356602987561745,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 18300
    },
    {
      "epoch": 4.356602987561745,
      "eval_runtime": 0.51,
      "eval_samples_per_second": 1.961,
      "eval_steps_per_second": 1.961,
      "step": 18300
    },
    {
      "epoch": 4.3568410402904245,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 18301
    },
    {
      "epoch": 4.357079093019104,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 18302
    },
    {
      "epoch": 4.357317145747783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 18303
    },
    {
      "epoch": 4.357555198476462,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 18304
    },
    {
      "epoch": 4.357793251205142,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 18305
    },
    {
      "epoch": 4.358031303933822,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 18306
    },
    {
      "epoch": 4.3582693566625,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 18307
    },
    {
      "epoch": 4.35850740939118,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 18308
    },
    {
      "epoch": 4.35874546211986,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 18309
    },
    {
      "epoch": 4.358983514848539,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 18310
    },
    {
      "epoch": 4.359221567577219,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3164,
      "step": 18311
    },
    {
      "epoch": 4.3594596203058975,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 18312
    },
    {
      "epoch": 4.359697673034577,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 18313
    },
    {
      "epoch": 4.359935725763257,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 18314
    },
    {
      "epoch": 4.360173778491936,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 18315
    },
    {
      "epoch": 4.360411831220615,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 18316
    },
    {
      "epoch": 4.360649883949295,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 18317
    },
    {
      "epoch": 4.360887936677974,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 18318
    },
    {
      "epoch": 4.361125989406654,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.379,
      "step": 18319
    },
    {
      "epoch": 4.361364042135333,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 18320
    },
    {
      "epoch": 4.361602094864012,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 18321
    },
    {
      "epoch": 4.361840147592692,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.4889,
      "step": 18322
    },
    {
      "epoch": 4.362078200321371,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 18323
    },
    {
      "epoch": 4.362316253050051,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 18324
    },
    {
      "epoch": 4.36255430577873,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 18325
    },
    {
      "epoch": 4.362792358507409,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 18326
    },
    {
      "epoch": 4.363030411236089,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 18327
    },
    {
      "epoch": 4.3632684639647685,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 18328
    },
    {
      "epoch": 4.363506516693447,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 18329
    },
    {
      "epoch": 4.363744569422127,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 18330
    },
    {
      "epoch": 4.3639826221508065,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 18331
    },
    {
      "epoch": 4.364220674879486,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 18332
    },
    {
      "epoch": 4.364458727608165,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 18333
    },
    {
      "epoch": 4.364696780336844,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 18334
    },
    {
      "epoch": 4.364934833065524,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 18335
    },
    {
      "epoch": 4.365172885794204,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 18336
    },
    {
      "epoch": 4.365410938522883,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 18337
    },
    {
      "epoch": 4.365648991251562,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 18338
    },
    {
      "epoch": 4.3658870439802415,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 18339
    },
    {
      "epoch": 4.366125096708921,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 18340
    },
    {
      "epoch": 4.366363149437601,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 18341
    },
    {
      "epoch": 4.3666012021662794,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4399,
      "step": 18342
    },
    {
      "epoch": 4.366839254894959,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 18343
    },
    {
      "epoch": 4.367077307623639,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3033,
      "step": 18344
    },
    {
      "epoch": 4.367315360352318,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 18345
    },
    {
      "epoch": 4.367553413080998,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 18346
    },
    {
      "epoch": 4.367791465809677,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 18347
    },
    {
      "epoch": 4.368029518538356,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 18348
    },
    {
      "epoch": 4.368267571267036,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 18349
    },
    {
      "epoch": 4.368505623995715,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 18350
    },
    {
      "epoch": 4.368743676724394,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 18351
    },
    {
      "epoch": 4.368981729453074,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2273,
      "step": 18352
    },
    {
      "epoch": 4.369219782181753,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 18353
    },
    {
      "epoch": 4.369457834910433,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 18354
    },
    {
      "epoch": 4.369695887639112,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 18355
    },
    {
      "epoch": 4.369933940367791,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 18356
    },
    {
      "epoch": 4.370171993096471,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 18357
    },
    {
      "epoch": 4.3704100458251505,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3102,
      "step": 18358
    },
    {
      "epoch": 4.37064809855383,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 18359
    },
    {
      "epoch": 4.370886151282509,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.237,
      "step": 18360
    },
    {
      "epoch": 4.371124204011188,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 18361
    },
    {
      "epoch": 4.371362256739868,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 18362
    },
    {
      "epoch": 4.371600309468548,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 18363
    },
    {
      "epoch": 4.371838362197226,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 18364
    },
    {
      "epoch": 4.372076414925906,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 18365
    },
    {
      "epoch": 4.3723144676545855,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 18366
    },
    {
      "epoch": 4.372552520383265,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 18367
    },
    {
      "epoch": 4.372790573111944,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 18368
    },
    {
      "epoch": 4.3730286258406235,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 18369
    },
    {
      "epoch": 4.373266678569303,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 18370
    },
    {
      "epoch": 4.373504731297983,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 18371
    },
    {
      "epoch": 4.373742784026662,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4012,
      "step": 18372
    },
    {
      "epoch": 4.373980836755341,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 18373
    },
    {
      "epoch": 4.374218889484021,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 18374
    },
    {
      "epoch": 4.3744569422127,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 18375
    },
    {
      "epoch": 4.37469499494138,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 18376
    },
    {
      "epoch": 4.3749330476700585,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 18377
    },
    {
      "epoch": 4.375171100398738,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 18378
    },
    {
      "epoch": 4.375409153127418,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 18379
    },
    {
      "epoch": 4.375647205856097,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 18380
    },
    {
      "epoch": 4.375885258584777,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 18381
    },
    {
      "epoch": 4.376123311313456,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 18382
    },
    {
      "epoch": 4.376361364042135,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 18383
    },
    {
      "epoch": 4.376599416770815,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 18384
    },
    {
      "epoch": 4.3768374694994945,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 18385
    },
    {
      "epoch": 4.377075522228173,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 18386
    },
    {
      "epoch": 4.377313574956853,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 18387
    },
    {
      "epoch": 4.377551627685532,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 18388
    },
    {
      "epoch": 4.377789680414212,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 18389
    },
    {
      "epoch": 4.378027733142892,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 18390
    },
    {
      "epoch": 4.37826578587157,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 18391
    },
    {
      "epoch": 4.37850383860025,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 18392
    },
    {
      "epoch": 4.3787418913289295,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 18393
    },
    {
      "epoch": 4.378979944057609,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 18394
    },
    {
      "epoch": 4.379217996786288,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 18395
    },
    {
      "epoch": 4.3794560495149675,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 18396
    },
    {
      "epoch": 4.379694102243647,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 18397
    },
    {
      "epoch": 4.379932154972327,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 18398
    },
    {
      "epoch": 4.380170207701005,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 18399
    },
    {
      "epoch": 4.380408260429685,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 18400
    },
    {
      "epoch": 4.380408260429685,
      "eval_runtime": 0.5317,
      "eval_samples_per_second": 1.881,
      "eval_steps_per_second": 1.881,
      "step": 18400
    },
    {
      "epoch": 4.380646313158365,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 18401
    },
    {
      "epoch": 4.380884365887044,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 18402
    },
    {
      "epoch": 4.381122418615723,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 18403
    },
    {
      "epoch": 4.3813604713444025,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.552,
      "step": 18404
    },
    {
      "epoch": 4.381598524073082,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3419,
      "step": 18405
    },
    {
      "epoch": 4.381836576801762,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3027,
      "step": 18406
    },
    {
      "epoch": 4.382074629530441,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 18407
    },
    {
      "epoch": 4.38231268225912,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 18408
    },
    {
      "epoch": 4.3825507349878,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 18409
    },
    {
      "epoch": 4.382788787716479,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 18410
    },
    {
      "epoch": 4.383026840445159,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 18411
    },
    {
      "epoch": 4.383264893173838,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 18412
    },
    {
      "epoch": 4.383502945902517,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 18413
    },
    {
      "epoch": 4.383740998631197,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 18414
    },
    {
      "epoch": 4.383979051359876,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 18415
    },
    {
      "epoch": 4.384217104088556,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 18416
    },
    {
      "epoch": 4.384455156817235,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 18417
    },
    {
      "epoch": 4.384693209545914,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 18418
    },
    {
      "epoch": 4.384931262274594,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 18419
    },
    {
      "epoch": 4.385169315003274,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 18420
    },
    {
      "epoch": 4.385407367731952,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 18421
    },
    {
      "epoch": 4.385645420460632,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 18422
    },
    {
      "epoch": 4.3858834731893115,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 18423
    },
    {
      "epoch": 4.386121525917991,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 18424
    },
    {
      "epoch": 4.386359578646671,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 18425
    },
    {
      "epoch": 4.386597631375349,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 18426
    },
    {
      "epoch": 4.386835684104029,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 18427
    },
    {
      "epoch": 4.387073736832709,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 18428
    },
    {
      "epoch": 4.387311789561388,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.5468,
      "step": 18429
    },
    {
      "epoch": 4.387549842290067,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2792,
      "step": 18430
    },
    {
      "epoch": 4.3877878950187466,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 18431
    },
    {
      "epoch": 4.388025947747426,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 18432
    },
    {
      "epoch": 4.388264000476106,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 18433
    },
    {
      "epoch": 4.3885020532047845,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 18434
    },
    {
      "epoch": 4.388740105933464,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 18435
    },
    {
      "epoch": 4.388978158662144,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 18436
    },
    {
      "epoch": 4.389216211390823,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 18437
    },
    {
      "epoch": 4.389454264119502,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 18438
    },
    {
      "epoch": 4.389692316848182,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 18439
    },
    {
      "epoch": 4.389930369576861,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 18440
    },
    {
      "epoch": 4.390168422305541,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 18441
    },
    {
      "epoch": 4.39040647503422,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 18442
    },
    {
      "epoch": 4.390644527762899,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 18443
    },
    {
      "epoch": 4.390882580491579,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 18444
    },
    {
      "epoch": 4.391120633220258,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 18445
    },
    {
      "epoch": 4.391358685948938,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 18446
    },
    {
      "epoch": 4.391596738677617,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 18447
    },
    {
      "epoch": 4.391834791406296,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.315,
      "step": 18448
    },
    {
      "epoch": 4.392072844134976,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.8835,
      "step": 18449
    },
    {
      "epoch": 4.3923108968636555,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 18450
    },
    {
      "epoch": 4.392548949592335,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 18451
    },
    {
      "epoch": 4.392787002321014,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 18452
    },
    {
      "epoch": 4.393025055049693,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 18453
    },
    {
      "epoch": 4.393263107778373,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 18454
    },
    {
      "epoch": 4.393501160507053,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 18455
    },
    {
      "epoch": 4.393739213235731,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 18456
    },
    {
      "epoch": 4.393977265964411,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 18457
    },
    {
      "epoch": 4.394215318693091,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 18458
    },
    {
      "epoch": 4.39445337142177,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 18459
    },
    {
      "epoch": 4.39469142415045,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 18460
    },
    {
      "epoch": 4.3949294768791285,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 18461
    },
    {
      "epoch": 4.395167529607808,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 18462
    },
    {
      "epoch": 4.395405582336488,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.4088,
      "step": 18463
    },
    {
      "epoch": 4.395643635065167,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 18464
    },
    {
      "epoch": 4.395881687793846,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 18465
    },
    {
      "epoch": 4.396119740522526,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 18466
    },
    {
      "epoch": 4.396357793251205,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 18467
    },
    {
      "epoch": 4.396595845979885,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 18468
    },
    {
      "epoch": 4.396833898708564,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 18469
    },
    {
      "epoch": 4.397071951437243,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.49,
      "step": 18470
    },
    {
      "epoch": 4.397310004165923,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 18471
    },
    {
      "epoch": 4.397548056894602,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 18472
    },
    {
      "epoch": 4.397786109623281,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 18473
    },
    {
      "epoch": 4.398024162351961,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 18474
    },
    {
      "epoch": 4.39826221508064,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3932,
      "step": 18475
    },
    {
      "epoch": 4.39850026780932,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 18476
    },
    {
      "epoch": 4.3987383205379995,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 18477
    },
    {
      "epoch": 4.398976373266678,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3679,
      "step": 18478
    },
    {
      "epoch": 4.399214425995358,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 18479
    },
    {
      "epoch": 4.399452478724037,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 18480
    },
    {
      "epoch": 4.399690531452717,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 18481
    },
    {
      "epoch": 4.399928584181396,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 18482
    },
    {
      "epoch": 4.400166636910075,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 18483
    },
    {
      "epoch": 4.400404689638755,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 18484
    },
    {
      "epoch": 4.400642742367435,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 18485
    },
    {
      "epoch": 4.400880795096114,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7398,
      "step": 18486
    },
    {
      "epoch": 4.401118847824793,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 18487
    },
    {
      "epoch": 4.4013569005534725,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 18488
    },
    {
      "epoch": 4.401594953282152,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2939,
      "step": 18489
    },
    {
      "epoch": 4.401833006010832,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 18490
    },
    {
      "epoch": 4.40207105873951,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4878,
      "step": 18491
    },
    {
      "epoch": 4.40230911146819,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 18492
    },
    {
      "epoch": 4.40254716419687,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 18493
    },
    {
      "epoch": 4.402785216925549,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 18494
    },
    {
      "epoch": 4.403023269654229,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 18495
    },
    {
      "epoch": 4.403261322382908,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 18496
    },
    {
      "epoch": 4.403499375111587,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 18497
    },
    {
      "epoch": 4.403737427840267,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 18498
    },
    {
      "epoch": 4.403975480568946,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 18499
    },
    {
      "epoch": 4.404213533297625,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 18500
    },
    {
      "epoch": 4.404213533297625,
      "eval_runtime": 0.5159,
      "eval_samples_per_second": 1.938,
      "eval_steps_per_second": 1.938,
      "step": 18500
    },
    {
      "epoch": 4.404451586026305,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 18501
    },
    {
      "epoch": 4.404689638754984,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.6458,
      "step": 18502
    },
    {
      "epoch": 4.404927691483664,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 18503
    },
    {
      "epoch": 4.405165744212343,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 18504
    },
    {
      "epoch": 4.405403796941022,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 18505
    },
    {
      "epoch": 4.405641849669702,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 18506
    },
    {
      "epoch": 4.4058799023983815,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3257,
      "step": 18507
    },
    {
      "epoch": 4.406117955127061,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 18508
    },
    {
      "epoch": 4.40635600785574,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 18509
    },
    {
      "epoch": 4.406594060584419,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 18510
    },
    {
      "epoch": 4.406832113313099,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 18511
    },
    {
      "epoch": 4.407070166041779,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 18512
    },
    {
      "epoch": 4.407308218770457,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 18513
    },
    {
      "epoch": 4.407546271499137,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.6892,
      "step": 18514
    },
    {
      "epoch": 4.4077843242278165,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 18515
    },
    {
      "epoch": 4.408022376956496,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 18516
    },
    {
      "epoch": 4.408260429685175,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 18517
    },
    {
      "epoch": 4.4084984824138544,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7324,
      "step": 18518
    },
    {
      "epoch": 4.408736535142534,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 18519
    },
    {
      "epoch": 4.408974587871214,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 18520
    },
    {
      "epoch": 4.409212640599893,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 18521
    },
    {
      "epoch": 4.409450693328572,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 18522
    },
    {
      "epoch": 4.409688746057252,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 18523
    },
    {
      "epoch": 4.409926798785931,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 18524
    },
    {
      "epoch": 4.410164851514611,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.4118,
      "step": 18525
    },
    {
      "epoch": 4.4104029042432895,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 18526
    },
    {
      "epoch": 4.410640956971969,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 18527
    },
    {
      "epoch": 4.410879009700649,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 18528
    },
    {
      "epoch": 4.411117062429328,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 18529
    },
    {
      "epoch": 4.411355115158008,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3947,
      "step": 18530
    },
    {
      "epoch": 4.411593167886687,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 18531
    },
    {
      "epoch": 4.411831220615366,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 18532
    },
    {
      "epoch": 4.412069273344046,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 18533
    },
    {
      "epoch": 4.4123073260727255,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3409,
      "step": 18534
    },
    {
      "epoch": 4.412545378801404,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 18535
    },
    {
      "epoch": 4.412783431530084,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 18536
    },
    {
      "epoch": 4.413021484258763,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3847,
      "step": 18537
    },
    {
      "epoch": 4.413259536987443,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 18538
    },
    {
      "epoch": 4.413497589716122,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 18539
    },
    {
      "epoch": 4.413735642444801,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 18540
    },
    {
      "epoch": 4.413973695173481,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 18541
    },
    {
      "epoch": 4.4142117479021605,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 18542
    },
    {
      "epoch": 4.41444980063084,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.4667,
      "step": 18543
    },
    {
      "epoch": 4.414687853359519,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 18544
    },
    {
      "epoch": 4.4149259060881985,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2087,
      "step": 18545
    },
    {
      "epoch": 4.415163958816878,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 18546
    },
    {
      "epoch": 4.415402011545558,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4905,
      "step": 18547
    },
    {
      "epoch": 4.415640064274236,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 18548
    },
    {
      "epoch": 4.415878117002916,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3806,
      "step": 18549
    },
    {
      "epoch": 4.416116169731596,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 18550
    },
    {
      "epoch": 4.416354222460275,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 18551
    },
    {
      "epoch": 4.416592275188954,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 18552
    },
    {
      "epoch": 4.4168303279176335,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.39,
      "step": 18553
    },
    {
      "epoch": 4.417068380646313,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 18554
    },
    {
      "epoch": 4.417306433374993,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2258,
      "step": 18555
    },
    {
      "epoch": 4.417544486103672,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4228,
      "step": 18556
    },
    {
      "epoch": 4.417782538832351,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 18557
    },
    {
      "epoch": 4.418020591561031,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 18558
    },
    {
      "epoch": 4.41825864428971,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 18559
    },
    {
      "epoch": 4.41849669701839,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3994,
      "step": 18560
    },
    {
      "epoch": 4.418734749747069,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 18561
    },
    {
      "epoch": 4.418972802475748,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 18562
    },
    {
      "epoch": 4.419210855204428,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.404,
      "step": 18563
    },
    {
      "epoch": 4.419448907933107,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 18564
    },
    {
      "epoch": 4.419686960661787,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 18565
    },
    {
      "epoch": 4.419925013390466,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 18566
    },
    {
      "epoch": 4.420163066119145,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 18567
    },
    {
      "epoch": 4.420401118847825,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 18568
    },
    {
      "epoch": 4.4206391715765045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 18569
    },
    {
      "epoch": 4.420877224305183,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 18570
    },
    {
      "epoch": 4.421115277033863,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 18571
    },
    {
      "epoch": 4.4213533297625425,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2965,
      "step": 18572
    },
    {
      "epoch": 4.421591382491222,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 18573
    },
    {
      "epoch": 4.421829435219901,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 18574
    },
    {
      "epoch": 4.42206748794858,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 18575
    },
    {
      "epoch": 4.42230554067726,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 18576
    },
    {
      "epoch": 4.42254359340594,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 18577
    },
    {
      "epoch": 4.422781646134619,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 18578
    },
    {
      "epoch": 4.423019698863298,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 18579
    },
    {
      "epoch": 4.4232577515919775,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 18580
    },
    {
      "epoch": 4.423495804320657,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2155,
      "step": 18581
    },
    {
      "epoch": 4.423733857049337,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 18582
    },
    {
      "epoch": 4.4239719097780155,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4613,
      "step": 18583
    },
    {
      "epoch": 4.424209962506695,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 18584
    },
    {
      "epoch": 4.424448015235375,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 18585
    },
    {
      "epoch": 4.424686067964054,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 18586
    },
    {
      "epoch": 4.424924120692733,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 18587
    },
    {
      "epoch": 4.425162173421413,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 18588
    },
    {
      "epoch": 4.425400226150092,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 18589
    },
    {
      "epoch": 4.425638278878772,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3555,
      "step": 18590
    },
    {
      "epoch": 4.425876331607451,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 18591
    },
    {
      "epoch": 4.42611438433613,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 18592
    },
    {
      "epoch": 4.42635243706481,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2893,
      "step": 18593
    },
    {
      "epoch": 4.426590489793489,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 18594
    },
    {
      "epoch": 4.426828542522169,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 18595
    },
    {
      "epoch": 4.427066595250848,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 18596
    },
    {
      "epoch": 4.427304647979527,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 18597
    },
    {
      "epoch": 4.427542700708207,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 18598
    },
    {
      "epoch": 4.4277807534368865,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 18599
    },
    {
      "epoch": 4.428018806165566,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 18600
    },
    {
      "epoch": 4.428018806165566,
      "eval_runtime": 0.5375,
      "eval_samples_per_second": 1.861,
      "eval_steps_per_second": 1.861,
      "step": 18600
    },
    {
      "epoch": 4.428256858894245,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 18601
    },
    {
      "epoch": 4.428494911622924,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 18602
    },
    {
      "epoch": 4.428732964351604,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 18603
    },
    {
      "epoch": 4.428971017080284,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 18604
    },
    {
      "epoch": 4.429209069808962,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 18605
    },
    {
      "epoch": 4.429447122537642,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 18606
    },
    {
      "epoch": 4.4296851752663216,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 18607
    },
    {
      "epoch": 4.429923227995001,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.237,
      "step": 18608
    },
    {
      "epoch": 4.430161280723681,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 18609
    },
    {
      "epoch": 4.4303993334523595,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 18610
    },
    {
      "epoch": 4.430637386181039,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 18611
    },
    {
      "epoch": 4.430875438909719,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 18612
    },
    {
      "epoch": 4.431113491638398,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 18613
    },
    {
      "epoch": 4.431351544367077,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 18614
    },
    {
      "epoch": 4.431589597095757,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 18615
    },
    {
      "epoch": 4.431827649824436,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 18616
    },
    {
      "epoch": 4.432065702553116,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 18617
    },
    {
      "epoch": 4.4323037552817945,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 18618
    },
    {
      "epoch": 4.432541808010474,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7669,
      "step": 18619
    },
    {
      "epoch": 4.432779860739154,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 18620
    },
    {
      "epoch": 4.433017913467833,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 18621
    },
    {
      "epoch": 4.433255966196512,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 18622
    },
    {
      "epoch": 4.433494018925192,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 18623
    },
    {
      "epoch": 4.433732071653871,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 18624
    },
    {
      "epoch": 4.433970124382551,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 18625
    },
    {
      "epoch": 4.4342081771112305,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 18626
    },
    {
      "epoch": 4.434446229839909,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 18627
    },
    {
      "epoch": 4.434684282568589,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 18628
    },
    {
      "epoch": 4.434922335297268,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 18629
    },
    {
      "epoch": 4.435160388025948,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 18630
    },
    {
      "epoch": 4.435398440754627,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 18631
    },
    {
      "epoch": 4.435636493483306,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 18632
    },
    {
      "epoch": 4.435874546211986,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 18633
    },
    {
      "epoch": 4.436112598940666,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 18634
    },
    {
      "epoch": 4.436350651669345,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 18635
    },
    {
      "epoch": 4.436588704398024,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 18636
    },
    {
      "epoch": 4.4368267571267035,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3591,
      "step": 18637
    },
    {
      "epoch": 4.437064809855383,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 18638
    },
    {
      "epoch": 4.437302862584063,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 18639
    },
    {
      "epoch": 4.437540915312741,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 18640
    },
    {
      "epoch": 4.437778968041421,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.6434,
      "step": 18641
    },
    {
      "epoch": 4.438017020770101,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 18642
    },
    {
      "epoch": 4.43825507349878,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 18643
    },
    {
      "epoch": 4.43849312622746,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8634,
      "step": 18644
    },
    {
      "epoch": 4.438731178956139,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 18645
    },
    {
      "epoch": 4.438969231684818,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 18646
    },
    {
      "epoch": 4.439207284413498,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 18647
    },
    {
      "epoch": 4.439445337142177,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 18648
    },
    {
      "epoch": 4.439683389870856,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 18649
    },
    {
      "epoch": 4.439921442599536,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 18650
    },
    {
      "epoch": 4.440159495328215,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 18651
    },
    {
      "epoch": 4.440397548056895,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 18652
    },
    {
      "epoch": 4.440635600785574,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 18653
    },
    {
      "epoch": 4.440873653514253,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 18654
    },
    {
      "epoch": 4.441111706242933,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 18655
    },
    {
      "epoch": 4.441349758971612,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8514,
      "step": 18656
    },
    {
      "epoch": 4.441587811700291,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.4789,
      "step": 18657
    },
    {
      "epoch": 4.441825864428971,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 18658
    },
    {
      "epoch": 4.44206391715765,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 18659
    },
    {
      "epoch": 4.44230196988633,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 18660
    },
    {
      "epoch": 4.44254002261501,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9181,
      "step": 18661
    },
    {
      "epoch": 4.442778075343688,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 18662
    },
    {
      "epoch": 4.443016128072368,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 18663
    },
    {
      "epoch": 4.4432541808010475,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 18664
    },
    {
      "epoch": 4.443492233529727,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3329,
      "step": 18665
    },
    {
      "epoch": 4.443730286258406,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 18666
    },
    {
      "epoch": 4.443968338987085,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 18667
    },
    {
      "epoch": 4.444206391715765,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 18668
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4847,
      "step": 18669
    },
    {
      "epoch": 4.444682497173124,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 18670
    },
    {
      "epoch": 4.444920549901803,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 18671
    },
    {
      "epoch": 4.445158602630483,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4402,
      "step": 18672
    },
    {
      "epoch": 4.445396655359162,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 18673
    },
    {
      "epoch": 4.445634708087842,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 18674
    },
    {
      "epoch": 4.4458727608165205,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 18675
    },
    {
      "epoch": 4.4461108135452,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 18676
    },
    {
      "epoch": 4.44634886627388,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 18677
    },
    {
      "epoch": 4.446586919002559,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3533,
      "step": 18678
    },
    {
      "epoch": 4.446824971731239,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 18679
    },
    {
      "epoch": 4.447063024459918,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 18680
    },
    {
      "epoch": 4.447301077188597,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 18681
    },
    {
      "epoch": 4.447539129917277,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 18682
    },
    {
      "epoch": 4.4477771826459565,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 18683
    },
    {
      "epoch": 4.448015235374635,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 18684
    },
    {
      "epoch": 4.448253288103315,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 18685
    },
    {
      "epoch": 4.448491340831994,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 18686
    },
    {
      "epoch": 4.448729393560674,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 18687
    },
    {
      "epoch": 4.448967446289353,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 18688
    },
    {
      "epoch": 4.449205499018032,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 18689
    },
    {
      "epoch": 4.449443551746712,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 18690
    },
    {
      "epoch": 4.4496816044753915,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 18691
    },
    {
      "epoch": 4.44991965720407,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 18692
    },
    {
      "epoch": 4.45015770993275,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 18693
    },
    {
      "epoch": 4.450395762661429,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 18694
    },
    {
      "epoch": 4.450633815390109,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 18695
    },
    {
      "epoch": 4.450871868118789,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 18696
    },
    {
      "epoch": 4.451109920847467,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 18697
    },
    {
      "epoch": 4.451347973576147,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 18698
    },
    {
      "epoch": 4.451586026304827,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 18699
    },
    {
      "epoch": 4.451824079033506,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 18700
    },
    {
      "epoch": 4.451824079033506,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 18700
    },
    {
      "epoch": 4.452062131762185,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 18701
    },
    {
      "epoch": 4.4523001844908645,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 18702
    },
    {
      "epoch": 4.452538237219544,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4634,
      "step": 18703
    },
    {
      "epoch": 4.452776289948224,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 18704
    },
    {
      "epoch": 4.453014342676903,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 18705
    },
    {
      "epoch": 4.453252395405582,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 18706
    },
    {
      "epoch": 4.453490448134262,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.358,
      "step": 18707
    },
    {
      "epoch": 4.453728500862941,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 18708
    },
    {
      "epoch": 4.453966553591621,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 18709
    },
    {
      "epoch": 4.4542046063203,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 18710
    },
    {
      "epoch": 4.454442659048979,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 18711
    },
    {
      "epoch": 4.454680711777659,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 18712
    },
    {
      "epoch": 4.454918764506338,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 18713
    },
    {
      "epoch": 4.455156817235018,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 18714
    },
    {
      "epoch": 4.455394869963697,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 18715
    },
    {
      "epoch": 4.455632922692376,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 18716
    },
    {
      "epoch": 4.455870975421056,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 18717
    },
    {
      "epoch": 4.4561090281497355,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 18718
    },
    {
      "epoch": 4.456347080878414,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 18719
    },
    {
      "epoch": 4.456585133607094,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 18720
    },
    {
      "epoch": 4.4568231863357735,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 18721
    },
    {
      "epoch": 4.457061239064453,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 18722
    },
    {
      "epoch": 4.457299291793132,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 18723
    },
    {
      "epoch": 4.457537344521811,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8314,
      "step": 18724
    },
    {
      "epoch": 4.457775397250491,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 18725
    },
    {
      "epoch": 4.458013449979171,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 18726
    },
    {
      "epoch": 4.45825150270785,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 18727
    },
    {
      "epoch": 4.458489555436529,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 18728
    },
    {
      "epoch": 4.4587276081652085,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 18729
    },
    {
      "epoch": 4.458965660893888,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 18730
    },
    {
      "epoch": 4.459203713622568,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 18731
    },
    {
      "epoch": 4.4594417663512465,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 18732
    },
    {
      "epoch": 4.459679819079926,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 18733
    },
    {
      "epoch": 4.459917871808606,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 18734
    },
    {
      "epoch": 4.460155924537285,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 18735
    },
    {
      "epoch": 4.460393977265964,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 18736
    },
    {
      "epoch": 4.460632029994644,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4022,
      "step": 18737
    },
    {
      "epoch": 4.460870082723323,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 18738
    },
    {
      "epoch": 4.461108135452003,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 18739
    },
    {
      "epoch": 4.461346188180682,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 18740
    },
    {
      "epoch": 4.461584240909361,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 18741
    },
    {
      "epoch": 4.461822293638041,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 18742
    },
    {
      "epoch": 4.46206034636672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 18743
    },
    {
      "epoch": 4.4622983990954,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 18744
    },
    {
      "epoch": 4.462536451824079,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 18745
    },
    {
      "epoch": 4.462774504552758,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 18746
    },
    {
      "epoch": 4.463012557281438,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 18747
    },
    {
      "epoch": 4.4632506100101175,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 18748
    },
    {
      "epoch": 4.463488662738797,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 18749
    },
    {
      "epoch": 4.463726715467476,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 18750
    },
    {
      "epoch": 4.463964768196155,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 18751
    },
    {
      "epoch": 4.464202820924835,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 18752
    },
    {
      "epoch": 4.464440873653515,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 18753
    },
    {
      "epoch": 4.464678926382193,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 18754
    },
    {
      "epoch": 4.464916979110873,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 18755
    },
    {
      "epoch": 4.4651550318395525,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4452,
      "step": 18756
    },
    {
      "epoch": 4.465393084568232,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 18757
    },
    {
      "epoch": 4.465631137296911,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 18758
    },
    {
      "epoch": 4.4658691900255905,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2879,
      "step": 18759
    },
    {
      "epoch": 4.46610724275427,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 18760
    },
    {
      "epoch": 4.46634529548295,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 18761
    },
    {
      "epoch": 4.466583348211629,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 18762
    },
    {
      "epoch": 4.466821400940308,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 18763
    },
    {
      "epoch": 4.467059453668988,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 18764
    },
    {
      "epoch": 4.467297506397667,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 18765
    },
    {
      "epoch": 4.467535559126347,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 18766
    },
    {
      "epoch": 4.4677736118550255,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 18767
    },
    {
      "epoch": 4.468011664583705,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 18768
    },
    {
      "epoch": 4.468249717312385,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5682,
      "step": 18769
    },
    {
      "epoch": 4.468487770041064,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 18770
    },
    {
      "epoch": 4.468725822769743,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 18771
    },
    {
      "epoch": 4.468963875498423,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 18772
    },
    {
      "epoch": 4.469201928227102,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 18773
    },
    {
      "epoch": 4.469439980955782,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 18774
    },
    {
      "epoch": 4.4696780336844615,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 18775
    },
    {
      "epoch": 4.46991608641314,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 18776
    },
    {
      "epoch": 4.47015413914182,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 18777
    },
    {
      "epoch": 4.470392191870499,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 18778
    },
    {
      "epoch": 4.470630244599179,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 18779
    },
    {
      "epoch": 4.470868297327858,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 18780
    },
    {
      "epoch": 4.471106350056537,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 18781
    },
    {
      "epoch": 4.471344402785217,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 18782
    },
    {
      "epoch": 4.4715824555138965,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 18783
    },
    {
      "epoch": 4.471820508242576,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 18784
    },
    {
      "epoch": 4.472058560971255,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 18785
    },
    {
      "epoch": 4.4722966136999345,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 18786
    },
    {
      "epoch": 4.472534666428614,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 18787
    },
    {
      "epoch": 4.472772719157294,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 18788
    },
    {
      "epoch": 4.473010771885972,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 18789
    },
    {
      "epoch": 4.473248824614652,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 18790
    },
    {
      "epoch": 4.473486877343332,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 18791
    },
    {
      "epoch": 4.473724930072011,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 18792
    },
    {
      "epoch": 4.473962982800691,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 18793
    },
    {
      "epoch": 4.4742010355293695,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 18794
    },
    {
      "epoch": 4.474439088258049,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 18795
    },
    {
      "epoch": 4.474677140986729,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 18796
    },
    {
      "epoch": 4.474915193715408,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 18797
    },
    {
      "epoch": 4.475153246444087,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 18798
    },
    {
      "epoch": 4.475391299172767,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 18799
    },
    {
      "epoch": 4.475629351901446,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 18800
    },
    {
      "epoch": 4.475629351901446,
      "eval_runtime": 0.5891,
      "eval_samples_per_second": 1.697,
      "eval_steps_per_second": 1.697,
      "step": 18800
    },
    {
      "epoch": 4.475867404630126,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 18801
    },
    {
      "epoch": 4.476105457358805,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 18802
    },
    {
      "epoch": 4.476343510087484,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 18803
    },
    {
      "epoch": 4.476581562816164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3418,
      "step": 18804
    },
    {
      "epoch": 4.476819615544843,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 18805
    },
    {
      "epoch": 4.477057668273522,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3123,
      "step": 18806
    },
    {
      "epoch": 4.477295721002202,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 18807
    },
    {
      "epoch": 4.477533773730881,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 18808
    },
    {
      "epoch": 4.477771826459561,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 18809
    },
    {
      "epoch": 4.478009879188241,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3402,
      "step": 18810
    },
    {
      "epoch": 4.478247931916919,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 18811
    },
    {
      "epoch": 4.478485984645599,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 18812
    },
    {
      "epoch": 4.4787240373742785,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 18813
    },
    {
      "epoch": 4.478962090102958,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 18814
    },
    {
      "epoch": 4.479200142831637,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 18815
    },
    {
      "epoch": 4.479438195560316,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3207,
      "step": 18816
    },
    {
      "epoch": 4.479676248288996,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 18817
    },
    {
      "epoch": 4.479914301017676,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 18818
    },
    {
      "epoch": 4.480152353746355,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 18819
    },
    {
      "epoch": 4.480390406475034,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.467,
      "step": 18820
    },
    {
      "epoch": 4.4806284592037136,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 18821
    },
    {
      "epoch": 4.480866511932393,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 18822
    },
    {
      "epoch": 4.481104564661073,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 18823
    },
    {
      "epoch": 4.4813426173897515,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 18824
    },
    {
      "epoch": 4.481580670118431,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4854,
      "step": 18825
    },
    {
      "epoch": 4.481818722847111,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 18826
    },
    {
      "epoch": 4.48205677557579,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 18827
    },
    {
      "epoch": 4.48229482830447,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 18828
    },
    {
      "epoch": 4.482532881033149,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 18829
    },
    {
      "epoch": 4.482770933761828,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 18830
    },
    {
      "epoch": 4.483008986490508,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 18831
    },
    {
      "epoch": 4.483247039219187,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 18832
    },
    {
      "epoch": 4.483485091947866,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 18833
    },
    {
      "epoch": 4.483723144676546,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 18834
    },
    {
      "epoch": 4.483961197405225,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 18835
    },
    {
      "epoch": 4.484199250133905,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 18836
    },
    {
      "epoch": 4.484437302862584,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 18837
    },
    {
      "epoch": 4.484675355591263,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 18838
    },
    {
      "epoch": 4.484913408319943,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 18839
    },
    {
      "epoch": 4.4851514610486225,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 18840
    },
    {
      "epoch": 4.485389513777301,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 18841
    },
    {
      "epoch": 4.485627566505981,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 18842
    },
    {
      "epoch": 4.48586561923466,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 18843
    },
    {
      "epoch": 4.48610367196334,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 18844
    },
    {
      "epoch": 4.48634172469202,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 18845
    },
    {
      "epoch": 4.486579777420698,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 18846
    },
    {
      "epoch": 4.486817830149378,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 18847
    },
    {
      "epoch": 4.487055882878058,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 18848
    },
    {
      "epoch": 4.487293935606737,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 18849
    },
    {
      "epoch": 4.487531988335416,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 18850
    },
    {
      "epoch": 4.4877700410640955,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 18851
    },
    {
      "epoch": 4.488008093792775,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 18852
    },
    {
      "epoch": 4.488246146521455,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 18853
    },
    {
      "epoch": 4.488484199250134,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 18854
    },
    {
      "epoch": 4.488722251978813,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.5736,
      "step": 18855
    },
    {
      "epoch": 4.488960304707493,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 18856
    },
    {
      "epoch": 4.489198357436172,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 18857
    },
    {
      "epoch": 4.489436410164852,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7473,
      "step": 18858
    },
    {
      "epoch": 4.489674462893531,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 18859
    },
    {
      "epoch": 4.48991251562221,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4425,
      "step": 18860
    },
    {
      "epoch": 4.49015056835089,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 18861
    },
    {
      "epoch": 4.490388621079569,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 18862
    },
    {
      "epoch": 4.490626673808249,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4345,
      "step": 18863
    },
    {
      "epoch": 4.490864726536928,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 18864
    },
    {
      "epoch": 4.491102779265607,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 18865
    },
    {
      "epoch": 4.491340831994287,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 18866
    },
    {
      "epoch": 4.4915788847229665,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 18867
    },
    {
      "epoch": 4.491816937451645,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 18868
    },
    {
      "epoch": 4.492054990180325,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5892,
      "step": 18869
    },
    {
      "epoch": 4.492293042909004,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 18870
    },
    {
      "epoch": 4.492531095637684,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8217,
      "step": 18871
    },
    {
      "epoch": 4.492769148366363,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 18872
    },
    {
      "epoch": 4.493007201095042,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 18873
    },
    {
      "epoch": 4.493245253823722,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 18874
    },
    {
      "epoch": 4.493483306552402,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4094,
      "step": 18875
    },
    {
      "epoch": 4.49372135928108,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 18876
    },
    {
      "epoch": 4.49395941200976,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 18877
    },
    {
      "epoch": 4.4941974647384395,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 18878
    },
    {
      "epoch": 4.494435517467119,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.6825,
      "step": 18879
    },
    {
      "epoch": 4.494673570195799,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 18880
    },
    {
      "epoch": 4.494911622924477,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 18881
    },
    {
      "epoch": 4.495149675653157,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 18882
    },
    {
      "epoch": 4.495387728381837,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 18883
    },
    {
      "epoch": 4.495625781110516,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 18884
    },
    {
      "epoch": 4.495863833839195,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 18885
    },
    {
      "epoch": 4.496101886567875,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 18886
    },
    {
      "epoch": 4.496339939296554,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 18887
    },
    {
      "epoch": 4.496577992025234,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 18888
    },
    {
      "epoch": 4.496816044753913,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 18889
    },
    {
      "epoch": 4.497054097482592,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 18890
    },
    {
      "epoch": 4.497292150211272,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 18891
    },
    {
      "epoch": 4.497530202939951,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 18892
    },
    {
      "epoch": 4.497768255668631,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 18893
    },
    {
      "epoch": 4.49800630839731,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 18894
    },
    {
      "epoch": 4.498244361125989,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 18895
    },
    {
      "epoch": 4.498482413854669,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 18896
    },
    {
      "epoch": 4.4987204665833485,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 18897
    },
    {
      "epoch": 4.498958519312028,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 18898
    },
    {
      "epoch": 4.499196572040707,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.15,
      "step": 18899
    },
    {
      "epoch": 4.499434624769386,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 18900
    },
    {
      "epoch": 4.499434624769386,
      "eval_runtime": 0.5183,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 18900
    },
    {
      "epoch": 4.499672677498066,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 18901
    },
    {
      "epoch": 4.499910730226746,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 18902
    },
    {
      "epoch": 4.500148782955424,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 18903
    },
    {
      "epoch": 4.500386835684104,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 18904
    },
    {
      "epoch": 4.5006248884127835,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 18905
    },
    {
      "epoch": 4.500862941141463,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 18906
    },
    {
      "epoch": 4.501100993870143,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 18907
    },
    {
      "epoch": 4.5013390465988214,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3438,
      "step": 18908
    },
    {
      "epoch": 4.501577099327501,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 18909
    },
    {
      "epoch": 4.501815152056181,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3371,
      "step": 18910
    },
    {
      "epoch": 4.502053204784859,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 18911
    },
    {
      "epoch": 4.502291257513539,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 18912
    },
    {
      "epoch": 4.502529310242219,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 18913
    },
    {
      "epoch": 4.502767362970898,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 18914
    },
    {
      "epoch": 4.503005415699578,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2577,
      "step": 18915
    },
    {
      "epoch": 4.5032434684282565,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 18916
    },
    {
      "epoch": 4.503481521156936,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 18917
    },
    {
      "epoch": 4.503719573885616,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 18918
    },
    {
      "epoch": 4.503957626614295,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 18919
    },
    {
      "epoch": 4.504195679342974,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3309,
      "step": 18920
    },
    {
      "epoch": 4.504433732071654,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 18921
    },
    {
      "epoch": 4.504671784800333,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 18922
    },
    {
      "epoch": 4.504909837529013,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 18923
    },
    {
      "epoch": 4.5051478902576925,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 18924
    },
    {
      "epoch": 4.505385942986371,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 18925
    },
    {
      "epoch": 4.505623995715051,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 18926
    },
    {
      "epoch": 4.50586204844373,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 18927
    },
    {
      "epoch": 4.50610010117241,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 18928
    },
    {
      "epoch": 4.506338153901089,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 18929
    },
    {
      "epoch": 4.506576206629768,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 18930
    },
    {
      "epoch": 4.506814259358448,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 18931
    },
    {
      "epoch": 4.5070523120871275,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 18932
    },
    {
      "epoch": 4.507290364815807,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 18933
    },
    {
      "epoch": 4.507528417544486,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 18934
    },
    {
      "epoch": 4.5077664702731655,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 18935
    },
    {
      "epoch": 4.508004523001845,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 18936
    },
    {
      "epoch": 4.508242575730525,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 18937
    },
    {
      "epoch": 4.508480628459203,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 18938
    },
    {
      "epoch": 4.508718681187883,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 18939
    },
    {
      "epoch": 4.508956733916563,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 18940
    },
    {
      "epoch": 4.509194786645242,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 18941
    },
    {
      "epoch": 4.509432839373922,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 18942
    },
    {
      "epoch": 4.5096708921026005,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 18943
    },
    {
      "epoch": 4.50990894483128,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 18944
    },
    {
      "epoch": 4.51014699755996,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 18945
    },
    {
      "epoch": 4.5103850502886385,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 18946
    },
    {
      "epoch": 4.510623103017318,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 18947
    },
    {
      "epoch": 4.510861155745998,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 18948
    },
    {
      "epoch": 4.511099208474677,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 18949
    },
    {
      "epoch": 4.511337261203357,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 18950
    },
    {
      "epoch": 4.511575313932036,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 18951
    },
    {
      "epoch": 4.511813366660715,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3268,
      "step": 18952
    },
    {
      "epoch": 4.512051419389395,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 18953
    },
    {
      "epoch": 4.512289472118074,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3712,
      "step": 18954
    },
    {
      "epoch": 4.512527524846753,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 18955
    },
    {
      "epoch": 4.512765577575433,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 18956
    },
    {
      "epoch": 4.513003630304112,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 18957
    },
    {
      "epoch": 4.513241683032792,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 18958
    },
    {
      "epoch": 4.5134797357614715,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 18959
    },
    {
      "epoch": 4.51371778849015,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 18960
    },
    {
      "epoch": 4.51395584121883,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 18961
    },
    {
      "epoch": 4.5141938939475095,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 18962
    },
    {
      "epoch": 4.514431946676189,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 18963
    },
    {
      "epoch": 4.514669999404868,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 18964
    },
    {
      "epoch": 4.514908052133547,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 18965
    },
    {
      "epoch": 4.515146104862227,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 18966
    },
    {
      "epoch": 4.515384157590907,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 18967
    },
    {
      "epoch": 4.515622210319586,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2004,
      "step": 18968
    },
    {
      "epoch": 4.515860263048265,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 18969
    },
    {
      "epoch": 4.5160983157769445,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3636,
      "step": 18970
    },
    {
      "epoch": 4.516336368505624,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 18971
    },
    {
      "epoch": 4.516574421234304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 18972
    },
    {
      "epoch": 4.5168124739629825,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 18973
    },
    {
      "epoch": 4.517050526691662,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 18974
    },
    {
      "epoch": 4.517288579420342,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 18975
    },
    {
      "epoch": 4.517526632149021,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 18976
    },
    {
      "epoch": 4.517764684877701,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 18977
    },
    {
      "epoch": 4.51800273760638,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 18978
    },
    {
      "epoch": 4.518240790335059,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 18979
    },
    {
      "epoch": 4.518478843063739,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4329,
      "step": 18980
    },
    {
      "epoch": 4.5187168957924175,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 18981
    },
    {
      "epoch": 4.518954948521097,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 18982
    },
    {
      "epoch": 4.519193001249777,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.2723,
      "step": 18983
    },
    {
      "epoch": 4.519431053978456,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 18984
    },
    {
      "epoch": 4.519669106707136,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 18985
    },
    {
      "epoch": 4.519907159435815,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 18986
    },
    {
      "epoch": 4.520145212164494,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 18987
    },
    {
      "epoch": 4.520383264893174,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 18988
    },
    {
      "epoch": 4.5206213176218535,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3834,
      "step": 18989
    },
    {
      "epoch": 4.520859370350532,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 18990
    },
    {
      "epoch": 4.521097423079212,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 18991
    },
    {
      "epoch": 4.521335475807891,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 18992
    },
    {
      "epoch": 4.521573528536571,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4384,
      "step": 18993
    },
    {
      "epoch": 4.521811581265251,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 18994
    },
    {
      "epoch": 4.522049633993929,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 18995
    },
    {
      "epoch": 4.522287686722609,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 18996
    },
    {
      "epoch": 4.5225257394512886,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 18997
    },
    {
      "epoch": 4.522763792179968,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 18998
    },
    {
      "epoch": 4.523001844908647,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 18999
    },
    {
      "epoch": 4.5232398976373265,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 19000
    },
    {
      "epoch": 4.5232398976373265,
      "eval_runtime": 0.5547,
      "eval_samples_per_second": 1.803,
      "eval_steps_per_second": 1.803,
      "step": 19000
    },
    {
      "epoch": 4.523477950366006,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 19001
    },
    {
      "epoch": 4.523716003094686,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 19002
    },
    {
      "epoch": 4.523954055823365,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 19003
    },
    {
      "epoch": 4.524192108552044,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 19004
    },
    {
      "epoch": 4.524430161280724,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 19005
    },
    {
      "epoch": 4.524668214009403,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 19006
    },
    {
      "epoch": 4.524906266738083,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 19007
    },
    {
      "epoch": 4.5251443194667615,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 19008
    },
    {
      "epoch": 4.525382372195441,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 19009
    },
    {
      "epoch": 4.525620424924121,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 19010
    },
    {
      "epoch": 4.5258584776528,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 19011
    },
    {
      "epoch": 4.52609653038148,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 19012
    },
    {
      "epoch": 4.526334583110159,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 19013
    },
    {
      "epoch": 4.526572635838838,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 19014
    },
    {
      "epoch": 4.526810688567518,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 19015
    },
    {
      "epoch": 4.527048741296197,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 19016
    },
    {
      "epoch": 4.527286794024876,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 19017
    },
    {
      "epoch": 4.527524846753556,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 19018
    },
    {
      "epoch": 4.527762899482235,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 19019
    },
    {
      "epoch": 4.528000952210915,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 19020
    },
    {
      "epoch": 4.528239004939594,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 19021
    },
    {
      "epoch": 4.528477057668273,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 19022
    },
    {
      "epoch": 4.528715110396953,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 19023
    },
    {
      "epoch": 4.528953163125633,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 19024
    },
    {
      "epoch": 4.529191215854311,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 19025
    },
    {
      "epoch": 4.529429268582991,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 19026
    },
    {
      "epoch": 4.5296673213116705,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 19027
    },
    {
      "epoch": 4.52990537404035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 19028
    },
    {
      "epoch": 4.53014342676903,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 19029
    },
    {
      "epoch": 4.530381479497708,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 19030
    },
    {
      "epoch": 4.530619532226388,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4063,
      "step": 19031
    },
    {
      "epoch": 4.530857584955068,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 19032
    },
    {
      "epoch": 4.531095637683747,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3325,
      "step": 19033
    },
    {
      "epoch": 4.531333690412426,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 19034
    },
    {
      "epoch": 4.531571743141106,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 19035
    },
    {
      "epoch": 4.531809795869785,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3167,
      "step": 19036
    },
    {
      "epoch": 4.532047848598465,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 19037
    },
    {
      "epoch": 4.532285901327144,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 19038
    },
    {
      "epoch": 4.532523954055823,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 19039
    },
    {
      "epoch": 4.532762006784503,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 19040
    },
    {
      "epoch": 4.533000059513182,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 19041
    },
    {
      "epoch": 4.533238112241862,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 19042
    },
    {
      "epoch": 4.533476164970541,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 19043
    },
    {
      "epoch": 4.53371421769922,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 19044
    },
    {
      "epoch": 4.5339522704279,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 19045
    },
    {
      "epoch": 4.534190323156579,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 19046
    },
    {
      "epoch": 4.534428375885259,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 19047
    },
    {
      "epoch": 4.534666428613938,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 19048
    },
    {
      "epoch": 4.534904481342617,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 19049
    },
    {
      "epoch": 4.535142534071297,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2789,
      "step": 19050
    },
    {
      "epoch": 4.535380586799977,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 19051
    },
    {
      "epoch": 4.535618639528655,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 19052
    },
    {
      "epoch": 4.535856692257335,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 19053
    },
    {
      "epoch": 4.5360947449860145,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 19054
    },
    {
      "epoch": 4.536332797714694,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 19055
    },
    {
      "epoch": 4.536570850443373,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 19056
    },
    {
      "epoch": 4.536808903172052,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 19057
    },
    {
      "epoch": 4.537046955900732,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 19058
    },
    {
      "epoch": 4.537285008629412,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 19059
    },
    {
      "epoch": 4.53752306135809,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 19060
    },
    {
      "epoch": 4.53776111408677,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 19061
    },
    {
      "epoch": 4.53799916681545,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 19062
    },
    {
      "epoch": 4.538237219544129,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 19063
    },
    {
      "epoch": 4.538475272272809,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.305,
      "step": 19064
    },
    {
      "epoch": 4.5387133250014875,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 19065
    },
    {
      "epoch": 4.538951377730167,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 19066
    },
    {
      "epoch": 4.539189430458847,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 19067
    },
    {
      "epoch": 4.539427483187526,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 19068
    },
    {
      "epoch": 4.539665535916205,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 19069
    },
    {
      "epoch": 4.539903588644885,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 19070
    },
    {
      "epoch": 4.540141641373564,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 19071
    },
    {
      "epoch": 4.540379694102244,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 19072
    },
    {
      "epoch": 4.5406177468309235,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 19073
    },
    {
      "epoch": 4.540855799559602,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 19074
    },
    {
      "epoch": 4.541093852288282,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 19075
    },
    {
      "epoch": 4.541331905016961,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 19076
    },
    {
      "epoch": 4.541569957745641,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 19077
    },
    {
      "epoch": 4.54180801047432,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 19078
    },
    {
      "epoch": 4.542046063202999,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 19079
    },
    {
      "epoch": 4.542284115931679,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 19080
    },
    {
      "epoch": 4.5425221686603585,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2899,
      "step": 19081
    },
    {
      "epoch": 4.542760221389038,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 19082
    },
    {
      "epoch": 4.542998274117717,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 19083
    },
    {
      "epoch": 4.5432363268463964,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 19084
    },
    {
      "epoch": 4.543474379575076,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 19085
    },
    {
      "epoch": 4.543712432303756,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 19086
    },
    {
      "epoch": 4.543950485032434,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 19087
    },
    {
      "epoch": 4.544188537761114,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 19088
    },
    {
      "epoch": 4.544426590489794,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 19089
    },
    {
      "epoch": 4.544664643218473,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 19090
    },
    {
      "epoch": 4.544902695947153,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 19091
    },
    {
      "epoch": 4.5451407486758315,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 19092
    },
    {
      "epoch": 4.545378801404511,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 19093
    },
    {
      "epoch": 4.545616854133191,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 19094
    },
    {
      "epoch": 4.545854906861869,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 19095
    },
    {
      "epoch": 4.546092959590549,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 19096
    },
    {
      "epoch": 4.546331012319229,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 19097
    },
    {
      "epoch": 4.546569065047908,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 19098
    },
    {
      "epoch": 4.546807117776588,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 19099
    },
    {
      "epoch": 4.547045170505267,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 19100
    },
    {
      "epoch": 4.547045170505267,
      "eval_runtime": 0.5236,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 19100
    },
    {
      "epoch": 4.547283223233946,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 19101
    },
    {
      "epoch": 4.547521275962626,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 19102
    },
    {
      "epoch": 4.547759328691305,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 19103
    },
    {
      "epoch": 4.547997381419984,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 19104
    },
    {
      "epoch": 4.548235434148664,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 19105
    },
    {
      "epoch": 4.548473486877343,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 19106
    },
    {
      "epoch": 4.548711539606023,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3458,
      "step": 19107
    },
    {
      "epoch": 4.5489495923347025,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 19108
    },
    {
      "epoch": 4.549187645063381,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 19109
    },
    {
      "epoch": 4.549425697792061,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 19110
    },
    {
      "epoch": 4.5496637505207405,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 19111
    },
    {
      "epoch": 4.54990180324942,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 19112
    },
    {
      "epoch": 4.550139855978099,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 19113
    },
    {
      "epoch": 4.550377908706778,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 19114
    },
    {
      "epoch": 4.550615961435458,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 19115
    },
    {
      "epoch": 4.550854014164138,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 19116
    },
    {
      "epoch": 4.551092066892817,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1574,
      "step": 19117
    },
    {
      "epoch": 4.551330119621496,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 19118
    },
    {
      "epoch": 4.5515681723501755,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 19119
    },
    {
      "epoch": 4.551806225078855,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 19120
    },
    {
      "epoch": 4.552044277807535,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4061,
      "step": 19121
    },
    {
      "epoch": 4.5522823305362135,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 19122
    },
    {
      "epoch": 4.552520383264893,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 19123
    },
    {
      "epoch": 4.552758435993573,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 19124
    },
    {
      "epoch": 4.552996488722252,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 19125
    },
    {
      "epoch": 4.553234541450932,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 19126
    },
    {
      "epoch": 4.553472594179611,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 19127
    },
    {
      "epoch": 4.55371064690829,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 19128
    },
    {
      "epoch": 4.55394869963697,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 19129
    },
    {
      "epoch": 4.5541867523656485,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 19130
    },
    {
      "epoch": 4.554424805094328,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 19131
    },
    {
      "epoch": 4.554662857823008,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 19132
    },
    {
      "epoch": 4.554900910551687,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 19133
    },
    {
      "epoch": 4.555138963280367,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 19134
    },
    {
      "epoch": 4.555377016009046,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 19135
    },
    {
      "epoch": 4.555615068737725,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 19136
    },
    {
      "epoch": 4.555853121466405,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 19137
    },
    {
      "epoch": 4.5560911741950845,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 19138
    },
    {
      "epoch": 4.556329226923763,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 19139
    },
    {
      "epoch": 4.556567279652443,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 19140
    },
    {
      "epoch": 4.556805332381122,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 19141
    },
    {
      "epoch": 4.557043385109802,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 19142
    },
    {
      "epoch": 4.557281437838482,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4129,
      "step": 19143
    },
    {
      "epoch": 4.55751949056716,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 19144
    },
    {
      "epoch": 4.55775754329584,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3614,
      "step": 19145
    },
    {
      "epoch": 4.5579955960245195,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 19146
    },
    {
      "epoch": 4.558233648753199,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 19147
    },
    {
      "epoch": 4.558471701481878,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 19148
    },
    {
      "epoch": 4.5587097542105575,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 19149
    },
    {
      "epoch": 4.558947806939237,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 19150
    },
    {
      "epoch": 4.559185859667917,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 19151
    },
    {
      "epoch": 4.559423912396596,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 19152
    },
    {
      "epoch": 4.559661965125275,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 19153
    },
    {
      "epoch": 4.559900017853955,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8092,
      "step": 19154
    },
    {
      "epoch": 4.560138070582634,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 19155
    },
    {
      "epoch": 4.560376123311314,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 19156
    },
    {
      "epoch": 4.5606141760399925,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 19157
    },
    {
      "epoch": 4.560852228768672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 19158
    },
    {
      "epoch": 4.561090281497352,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 19159
    },
    {
      "epoch": 4.561328334226031,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 19160
    },
    {
      "epoch": 4.561566386954711,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 19161
    },
    {
      "epoch": 4.56180443968339,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 19162
    },
    {
      "epoch": 4.562042492412069,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 19163
    },
    {
      "epoch": 4.562280545140749,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.5929,
      "step": 19164
    },
    {
      "epoch": 4.562518597869428,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 19165
    },
    {
      "epoch": 4.562756650598107,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 19166
    },
    {
      "epoch": 4.562994703326787,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 19167
    },
    {
      "epoch": 4.563232756055466,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 19168
    },
    {
      "epoch": 4.563470808784146,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 19169
    },
    {
      "epoch": 4.563708861512825,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 19170
    },
    {
      "epoch": 4.563946914241504,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 19171
    },
    {
      "epoch": 4.564184966970184,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 19172
    },
    {
      "epoch": 4.5644230196988635,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 19173
    },
    {
      "epoch": 4.564661072427542,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 19174
    },
    {
      "epoch": 4.564899125156222,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 19175
    },
    {
      "epoch": 4.5651371778849015,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.6917,
      "step": 19176
    },
    {
      "epoch": 4.565375230613581,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 19177
    },
    {
      "epoch": 4.565613283342261,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 19178
    },
    {
      "epoch": 4.565851336070939,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4038,
      "step": 19179
    },
    {
      "epoch": 4.566089388799619,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 19180
    },
    {
      "epoch": 4.566327441528299,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 19181
    },
    {
      "epoch": 4.566565494256978,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 19182
    },
    {
      "epoch": 4.566803546985657,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 19183
    },
    {
      "epoch": 4.5670415997143365,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 19184
    },
    {
      "epoch": 4.567279652443016,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 19185
    },
    {
      "epoch": 4.567517705171696,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 19186
    },
    {
      "epoch": 4.567755757900375,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3091,
      "step": 19187
    },
    {
      "epoch": 4.567993810629054,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 19188
    },
    {
      "epoch": 4.568231863357734,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 19189
    },
    {
      "epoch": 4.568469916086413,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 19190
    },
    {
      "epoch": 4.568707968815093,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 19191
    },
    {
      "epoch": 4.568946021543772,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 19192
    },
    {
      "epoch": 4.569184074272451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 19193
    },
    {
      "epoch": 4.569422127001131,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 19194
    },
    {
      "epoch": 4.56966017972981,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 19195
    },
    {
      "epoch": 4.56989823245849,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 19196
    },
    {
      "epoch": 4.570136285187169,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 19197
    },
    {
      "epoch": 4.570374337915848,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 19198
    },
    {
      "epoch": 4.570612390644528,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 19199
    },
    {
      "epoch": 4.570850443373207,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 19200
    },
    {
      "epoch": 4.570850443373207,
      "eval_runtime": 0.5257,
      "eval_samples_per_second": 1.902,
      "eval_steps_per_second": 1.902,
      "step": 19200
    },
    {
      "epoch": 4.571088496101886,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 19201
    },
    {
      "epoch": 4.571326548830566,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 19202
    },
    {
      "epoch": 4.5715646015592455,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 19203
    },
    {
      "epoch": 4.571802654287925,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 19204
    },
    {
      "epoch": 4.572040707016604,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 19205
    },
    {
      "epoch": 4.572278759745283,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 19206
    },
    {
      "epoch": 4.572516812473963,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 19207
    },
    {
      "epoch": 4.572754865202643,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 19208
    },
    {
      "epoch": 4.572992917931321,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5342,
      "step": 19209
    },
    {
      "epoch": 4.573230970660001,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3305,
      "step": 19210
    },
    {
      "epoch": 4.5734690233886806,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 19211
    },
    {
      "epoch": 4.57370707611736,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 19212
    },
    {
      "epoch": 4.57394512884604,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 19213
    },
    {
      "epoch": 4.5741831815747185,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 19214
    },
    {
      "epoch": 4.574421234303398,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 19215
    },
    {
      "epoch": 4.574659287032078,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 19216
    },
    {
      "epoch": 4.574897339760757,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 19217
    },
    {
      "epoch": 4.575135392489436,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 19218
    },
    {
      "epoch": 4.575373445218116,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 19219
    },
    {
      "epoch": 4.575611497946795,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 19220
    },
    {
      "epoch": 4.575849550675475,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 19221
    },
    {
      "epoch": 4.576087603404154,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 19222
    },
    {
      "epoch": 4.576325656132833,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 19223
    },
    {
      "epoch": 4.576563708861513,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 19224
    },
    {
      "epoch": 4.576801761590192,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 19225
    },
    {
      "epoch": 4.577039814318872,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 19226
    },
    {
      "epoch": 4.577277867047551,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 19227
    },
    {
      "epoch": 4.57751591977623,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 19228
    },
    {
      "epoch": 4.57775397250491,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8965,
      "step": 19229
    },
    {
      "epoch": 4.5779920252335895,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 19230
    },
    {
      "epoch": 4.578230077962269,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 19231
    },
    {
      "epoch": 4.578468130690948,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 19232
    },
    {
      "epoch": 4.578706183419627,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 19233
    },
    {
      "epoch": 4.578944236148307,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 19234
    },
    {
      "epoch": 4.579182288876987,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 19235
    },
    {
      "epoch": 4.579420341605665,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 19236
    },
    {
      "epoch": 4.579658394334345,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 19237
    },
    {
      "epoch": 4.579896447063025,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 19238
    },
    {
      "epoch": 4.580134499791704,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 19239
    },
    {
      "epoch": 4.580372552520383,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 19240
    },
    {
      "epoch": 4.5806106052490625,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 19241
    },
    {
      "epoch": 4.580848657977742,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 19242
    },
    {
      "epoch": 4.581086710706422,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 19243
    },
    {
      "epoch": 4.5813247634351,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 19244
    },
    {
      "epoch": 4.58156281616378,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 19245
    },
    {
      "epoch": 4.58180086889246,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 19246
    },
    {
      "epoch": 4.582038921621139,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 19247
    },
    {
      "epoch": 4.582276974349819,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 19248
    },
    {
      "epoch": 4.582515027078498,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 19249
    },
    {
      "epoch": 4.582753079807177,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 19250
    },
    {
      "epoch": 4.582991132535857,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 19251
    },
    {
      "epoch": 4.583229185264536,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 19252
    },
    {
      "epoch": 4.583467237993215,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 19253
    },
    {
      "epoch": 4.583705290721895,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 19254
    },
    {
      "epoch": 4.583943343450574,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 19255
    },
    {
      "epoch": 4.584181396179254,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 19256
    },
    {
      "epoch": 4.5844194489079335,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 19257
    },
    {
      "epoch": 4.584657501636612,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 19258
    },
    {
      "epoch": 4.584895554365292,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 19259
    },
    {
      "epoch": 4.585133607093971,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 19260
    },
    {
      "epoch": 4.585371659822651,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.5738,
      "step": 19261
    },
    {
      "epoch": 4.58560971255133,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 19262
    },
    {
      "epoch": 4.585847765280009,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 19263
    },
    {
      "epoch": 4.586085818008689,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 19264
    },
    {
      "epoch": 4.586323870737369,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 19265
    },
    {
      "epoch": 4.586561923466048,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 19266
    },
    {
      "epoch": 4.586799976194727,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 19267
    },
    {
      "epoch": 4.5870380289234065,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 19268
    },
    {
      "epoch": 4.587276081652086,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 19269
    },
    {
      "epoch": 4.587514134380766,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 19270
    },
    {
      "epoch": 4.587752187109444,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 19271
    },
    {
      "epoch": 4.587990239838124,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 19272
    },
    {
      "epoch": 4.588228292566804,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 19273
    },
    {
      "epoch": 4.588466345295483,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 19274
    },
    {
      "epoch": 4.588704398024162,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 19275
    },
    {
      "epoch": 4.588942450752842,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 19276
    },
    {
      "epoch": 4.589180503481521,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 19277
    },
    {
      "epoch": 4.589418556210201,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 19278
    },
    {
      "epoch": 4.5896566089388795,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 19279
    },
    {
      "epoch": 4.589894661667559,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 19280
    },
    {
      "epoch": 4.590132714396239,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 19281
    },
    {
      "epoch": 4.590370767124918,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 19282
    },
    {
      "epoch": 4.590608819853598,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 19283
    },
    {
      "epoch": 4.590846872582277,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 19284
    },
    {
      "epoch": 4.591084925310956,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 19285
    },
    {
      "epoch": 4.591322978039636,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 19286
    },
    {
      "epoch": 4.5915610307683155,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 19287
    },
    {
      "epoch": 4.591799083496994,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 19288
    },
    {
      "epoch": 4.592037136225674,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 19289
    },
    {
      "epoch": 4.592275188954353,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 19290
    },
    {
      "epoch": 4.592513241683033,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.7149,
      "step": 19291
    },
    {
      "epoch": 4.592751294411713,
      "grad_norm": 2.6875,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 19292
    },
    {
      "epoch": 4.592989347140391,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4564,
      "step": 19293
    },
    {
      "epoch": 4.593227399869071,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 19294
    },
    {
      "epoch": 4.5934654525977505,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 19295
    },
    {
      "epoch": 4.59370350532643,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 19296
    },
    {
      "epoch": 4.593941558055109,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 19297
    },
    {
      "epoch": 4.5941796107837884,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 19298
    },
    {
      "epoch": 4.594417663512468,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 19299
    },
    {
      "epoch": 4.594655716241148,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 19300
    },
    {
      "epoch": 4.594655716241148,
      "eval_runtime": 0.5323,
      "eval_samples_per_second": 1.879,
      "eval_steps_per_second": 1.879,
      "step": 19300
    },
    {
      "epoch": 4.594893768969827,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4232,
      "step": 19301
    },
    {
      "epoch": 4.595131821698506,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 19302
    },
    {
      "epoch": 4.595369874427186,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 19303
    },
    {
      "epoch": 4.595607927155865,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 19304
    },
    {
      "epoch": 4.595845979884545,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 19305
    },
    {
      "epoch": 4.5960840326132235,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 19306
    },
    {
      "epoch": 4.596322085341903,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 19307
    },
    {
      "epoch": 4.596560138070583,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3402,
      "step": 19308
    },
    {
      "epoch": 4.596798190799262,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 19309
    },
    {
      "epoch": 4.597036243527942,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 19310
    },
    {
      "epoch": 4.597274296256621,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 19311
    },
    {
      "epoch": 4.5975123489853,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 19312
    },
    {
      "epoch": 4.59775040171398,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 19313
    },
    {
      "epoch": 4.597988454442659,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 19314
    },
    {
      "epoch": 4.598226507171338,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 19315
    },
    {
      "epoch": 4.598464559900018,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 19316
    },
    {
      "epoch": 4.598702612628697,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 19317
    },
    {
      "epoch": 4.598940665357377,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 19318
    },
    {
      "epoch": 4.599178718086056,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 19319
    },
    {
      "epoch": 4.599416770814735,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 19320
    },
    {
      "epoch": 4.599654823543415,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 19321
    },
    {
      "epoch": 4.5998928762720945,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3386,
      "step": 19322
    },
    {
      "epoch": 4.600130929000773,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 19323
    },
    {
      "epoch": 4.600368981729453,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 19324
    },
    {
      "epoch": 4.6006070344581325,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 19325
    },
    {
      "epoch": 4.600845087186812,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 19326
    },
    {
      "epoch": 4.601083139915492,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 19327
    },
    {
      "epoch": 4.60132119264417,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 19328
    },
    {
      "epoch": 4.60155924537285,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 19329
    },
    {
      "epoch": 4.60179729810153,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 19330
    },
    {
      "epoch": 4.602035350830209,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 19331
    },
    {
      "epoch": 4.602273403558888,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 19332
    },
    {
      "epoch": 4.6025114562875675,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2365,
      "step": 19333
    },
    {
      "epoch": 4.602749509016247,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4865,
      "step": 19334
    },
    {
      "epoch": 4.602987561744927,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 19335
    },
    {
      "epoch": 4.603225614473606,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 19336
    },
    {
      "epoch": 4.603463667202285,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 19337
    },
    {
      "epoch": 4.603701719930965,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 19338
    },
    {
      "epoch": 4.603939772659644,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 19339
    },
    {
      "epoch": 4.604177825388324,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 19340
    },
    {
      "epoch": 4.604415878117003,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 19341
    },
    {
      "epoch": 4.604653930845682,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 19342
    },
    {
      "epoch": 4.604891983574362,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 19343
    },
    {
      "epoch": 4.605130036303041,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.5436,
      "step": 19344
    },
    {
      "epoch": 4.605368089031721,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 19345
    },
    {
      "epoch": 4.6056061417604,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 19346
    },
    {
      "epoch": 4.605844194489079,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 19347
    },
    {
      "epoch": 4.606082247217759,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 19348
    },
    {
      "epoch": 4.606320299946438,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 19349
    },
    {
      "epoch": 4.606558352675117,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 19350
    },
    {
      "epoch": 4.606796405403797,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 19351
    },
    {
      "epoch": 4.6070344581324765,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 19352
    },
    {
      "epoch": 4.607272510861156,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 19353
    },
    {
      "epoch": 4.607510563589835,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 19354
    },
    {
      "epoch": 4.607748616318514,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 19355
    },
    {
      "epoch": 4.607986669047194,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 19356
    },
    {
      "epoch": 4.608224721775874,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 19357
    },
    {
      "epoch": 4.608462774504552,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 19358
    },
    {
      "epoch": 4.608700827233232,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 19359
    },
    {
      "epoch": 4.6089388799619115,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 19360
    },
    {
      "epoch": 4.609176932690591,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 19361
    },
    {
      "epoch": 4.609414985419271,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 19362
    },
    {
      "epoch": 4.6096530381479495,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 19363
    },
    {
      "epoch": 4.609891090876629,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 19364
    },
    {
      "epoch": 4.610129143605309,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2896,
      "step": 19365
    },
    {
      "epoch": 4.610367196333988,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 19366
    },
    {
      "epoch": 4.610605249062667,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 19367
    },
    {
      "epoch": 4.610843301791347,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 19368
    },
    {
      "epoch": 4.611081354520026,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 19369
    },
    {
      "epoch": 4.611319407248706,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 19370
    },
    {
      "epoch": 4.611557459977385,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 19371
    },
    {
      "epoch": 4.611795512706064,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 19372
    },
    {
      "epoch": 4.612033565434744,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 19373
    },
    {
      "epoch": 4.612271618163423,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 19374
    },
    {
      "epoch": 4.612509670892103,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.8636,
      "step": 19375
    },
    {
      "epoch": 4.612747723620782,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 19376
    },
    {
      "epoch": 4.612985776349461,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 19377
    },
    {
      "epoch": 4.613223829078141,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 19378
    },
    {
      "epoch": 4.6134618818068205,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 19379
    },
    {
      "epoch": 4.6136999345355,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 19380
    },
    {
      "epoch": 4.613937987264179,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4711,
      "step": 19381
    },
    {
      "epoch": 4.614176039992858,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.442,
      "step": 19382
    },
    {
      "epoch": 4.614414092721538,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 19383
    },
    {
      "epoch": 4.614652145450217,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.4219,
      "step": 19384
    },
    {
      "epoch": 4.614890198178896,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 19385
    },
    {
      "epoch": 4.615128250907576,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 19386
    },
    {
      "epoch": 4.6153663036362556,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 19387
    },
    {
      "epoch": 4.615604356364935,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 19388
    },
    {
      "epoch": 4.615842409093614,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3669,
      "step": 19389
    },
    {
      "epoch": 4.6160804618222935,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 19390
    },
    {
      "epoch": 4.616318514550973,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3369,
      "step": 19391
    },
    {
      "epoch": 4.616556567279653,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 19392
    },
    {
      "epoch": 4.616794620008331,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 19393
    },
    {
      "epoch": 4.617032672737011,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 19394
    },
    {
      "epoch": 4.617270725465691,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 19395
    },
    {
      "epoch": 4.61750877819437,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 19396
    },
    {
      "epoch": 4.61774683092305,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 19397
    },
    {
      "epoch": 4.6179848836517285,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2922,
      "step": 19398
    },
    {
      "epoch": 4.618222936380408,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 19399
    },
    {
      "epoch": 4.618460989109088,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 19400
    },
    {
      "epoch": 4.618460989109088,
      "eval_runtime": 0.5133,
      "eval_samples_per_second": 1.948,
      "eval_steps_per_second": 1.948,
      "step": 19400
    },
    {
      "epoch": 4.618699041837767,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 19401
    },
    {
      "epoch": 4.618937094566446,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.4477,
      "step": 19402
    },
    {
      "epoch": 4.619175147295126,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 19403
    },
    {
      "epoch": 4.619413200023805,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 19404
    },
    {
      "epoch": 4.619651252752485,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 19405
    },
    {
      "epoch": 4.6198893054811645,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 19406
    },
    {
      "epoch": 4.620127358209843,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 19407
    },
    {
      "epoch": 4.620365410938523,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 19408
    },
    {
      "epoch": 4.620603463667202,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 19409
    },
    {
      "epoch": 4.620841516395882,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 19410
    },
    {
      "epoch": 4.621079569124561,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 19411
    },
    {
      "epoch": 4.62131762185324,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.6294,
      "step": 19412
    },
    {
      "epoch": 4.62155567458192,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 19413
    },
    {
      "epoch": 4.6217937273106,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 19414
    },
    {
      "epoch": 4.622031780039279,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 19415
    },
    {
      "epoch": 4.622269832767958,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 19416
    },
    {
      "epoch": 4.6225078854966375,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 19417
    },
    {
      "epoch": 4.622745938225317,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 19418
    },
    {
      "epoch": 4.622983990953996,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 19419
    },
    {
      "epoch": 4.623222043682675,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 19420
    },
    {
      "epoch": 4.623460096411355,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 19421
    },
    {
      "epoch": 4.623698149140035,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1513,
      "step": 19422
    },
    {
      "epoch": 4.623936201868714,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 19423
    },
    {
      "epoch": 4.624174254597393,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 19424
    },
    {
      "epoch": 4.624412307326073,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 19425
    },
    {
      "epoch": 4.624650360054752,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 19426
    },
    {
      "epoch": 4.624888412783432,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 19427
    },
    {
      "epoch": 4.6251264655121105,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 19428
    },
    {
      "epoch": 4.62536451824079,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 19429
    },
    {
      "epoch": 4.62560257096947,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 19430
    },
    {
      "epoch": 4.625840623698149,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 19431
    },
    {
      "epoch": 4.626078676426829,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 19432
    },
    {
      "epoch": 4.626316729155508,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 19433
    },
    {
      "epoch": 4.626554781884187,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 19434
    },
    {
      "epoch": 4.626792834612867,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 19435
    },
    {
      "epoch": 4.627030887341546,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 19436
    },
    {
      "epoch": 4.627268940070225,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 19437
    },
    {
      "epoch": 4.627506992798905,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 19438
    },
    {
      "epoch": 4.627745045527584,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 19439
    },
    {
      "epoch": 4.627983098256264,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 19440
    },
    {
      "epoch": 4.628221150984944,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 19441
    },
    {
      "epoch": 4.628459203713622,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 19442
    },
    {
      "epoch": 4.628697256442302,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2547,
      "step": 19443
    },
    {
      "epoch": 4.6289353091709815,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 19444
    },
    {
      "epoch": 4.629173361899661,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 19445
    },
    {
      "epoch": 4.62941141462834,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 19446
    },
    {
      "epoch": 4.629649467357019,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 19447
    },
    {
      "epoch": 4.629887520085699,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 19448
    },
    {
      "epoch": 4.630125572814379,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 19449
    },
    {
      "epoch": 4.630363625543058,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 19450
    },
    {
      "epoch": 4.630601678271737,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 19451
    },
    {
      "epoch": 4.630839731000417,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 19452
    },
    {
      "epoch": 4.631077783729096,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 19453
    },
    {
      "epoch": 4.631315836457776,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2234,
      "step": 19454
    },
    {
      "epoch": 4.6315538891864545,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4384,
      "step": 19455
    },
    {
      "epoch": 4.631791941915134,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 19456
    },
    {
      "epoch": 4.632029994643814,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 19457
    },
    {
      "epoch": 4.632268047372493,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.365,
      "step": 19458
    },
    {
      "epoch": 4.632506100101172,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 19459
    },
    {
      "epoch": 4.632744152829852,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 19460
    },
    {
      "epoch": 4.632982205558531,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 19461
    },
    {
      "epoch": 4.633220258287211,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 19462
    },
    {
      "epoch": 4.63345831101589,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 19463
    },
    {
      "epoch": 4.633696363744569,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 19464
    },
    {
      "epoch": 4.633934416473249,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 19465
    },
    {
      "epoch": 4.634172469201928,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 19466
    },
    {
      "epoch": 4.634410521930608,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 19467
    },
    {
      "epoch": 4.634648574659287,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 19468
    },
    {
      "epoch": 4.634886627387966,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2919,
      "step": 19469
    },
    {
      "epoch": 4.635124680116646,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 19470
    },
    {
      "epoch": 4.6353627328453255,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 19471
    },
    {
      "epoch": 4.635600785574004,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3893,
      "step": 19472
    },
    {
      "epoch": 4.635838838302684,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 19473
    },
    {
      "epoch": 4.6360768910313634,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 19474
    },
    {
      "epoch": 4.636314943760043,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 19475
    },
    {
      "epoch": 4.636552996488723,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 19476
    },
    {
      "epoch": 4.636791049217401,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3421,
      "step": 19477
    },
    {
      "epoch": 4.637029101946081,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 19478
    },
    {
      "epoch": 4.637267154674761,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 19479
    },
    {
      "epoch": 4.63750520740344,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 19480
    },
    {
      "epoch": 4.637743260132119,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 19481
    },
    {
      "epoch": 4.6379813128607985,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 19482
    },
    {
      "epoch": 4.638219365589478,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 19483
    },
    {
      "epoch": 4.638457418318158,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 19484
    },
    {
      "epoch": 4.638695471046837,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 19485
    },
    {
      "epoch": 4.638933523775516,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 19486
    },
    {
      "epoch": 4.639171576504196,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 19487
    },
    {
      "epoch": 4.639409629232875,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 19488
    },
    {
      "epoch": 4.639647681961555,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 19489
    },
    {
      "epoch": 4.639885734690234,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 19490
    },
    {
      "epoch": 4.640123787418913,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 19491
    },
    {
      "epoch": 4.640361840147593,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 19492
    },
    {
      "epoch": 4.640599892876272,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 19493
    },
    {
      "epoch": 4.640837945604951,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3218,
      "step": 19494
    },
    {
      "epoch": 4.641075998333631,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 19495
    },
    {
      "epoch": 4.64131405106231,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 19496
    },
    {
      "epoch": 4.64155210379099,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 19497
    },
    {
      "epoch": 4.641790156519669,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 19498
    },
    {
      "epoch": 4.642028209248348,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 19499
    },
    {
      "epoch": 4.642266261977028,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 19500
    },
    {
      "epoch": 4.642266261977028,
      "eval_runtime": 0.5185,
      "eval_samples_per_second": 1.928,
      "eval_steps_per_second": 1.928,
      "step": 19500
    },
    {
      "epoch": 4.6425043147057075,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 19501
    },
    {
      "epoch": 4.642742367434387,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 19502
    },
    {
      "epoch": 4.642980420163066,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 19503
    },
    {
      "epoch": 4.643218472891745,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 19504
    },
    {
      "epoch": 4.643456525620425,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 19505
    },
    {
      "epoch": 4.643694578349105,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3725,
      "step": 19506
    },
    {
      "epoch": 4.643932631077783,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 19507
    },
    {
      "epoch": 4.644170683806463,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 19508
    },
    {
      "epoch": 4.6444087365351425,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 19509
    },
    {
      "epoch": 4.644646789263822,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 19510
    },
    {
      "epoch": 4.644884841992502,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 19511
    },
    {
      "epoch": 4.6451228947211805,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 19512
    },
    {
      "epoch": 4.64536094744986,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 19513
    },
    {
      "epoch": 4.64559900017854,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 19514
    },
    {
      "epoch": 4.645837052907219,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 19515
    },
    {
      "epoch": 4.646075105635898,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 19516
    },
    {
      "epoch": 4.646313158364578,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 19517
    },
    {
      "epoch": 4.646551211093257,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 19518
    },
    {
      "epoch": 4.646789263821937,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 19519
    },
    {
      "epoch": 4.647027316550616,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 19520
    },
    {
      "epoch": 4.647265369279295,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 19521
    },
    {
      "epoch": 4.647503422007975,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 19522
    },
    {
      "epoch": 4.647741474736654,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 19523
    },
    {
      "epoch": 4.647979527465334,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3205,
      "step": 19524
    },
    {
      "epoch": 4.648217580194013,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 19525
    },
    {
      "epoch": 4.648455632922692,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 19526
    },
    {
      "epoch": 4.648693685651372,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8773,
      "step": 19527
    },
    {
      "epoch": 4.6489317383800515,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 19528
    },
    {
      "epoch": 4.649169791108731,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 19529
    },
    {
      "epoch": 4.64940784383741,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 19530
    },
    {
      "epoch": 4.649645896566089,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.256,
      "step": 19531
    },
    {
      "epoch": 4.649883949294769,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 19532
    },
    {
      "epoch": 4.650122002023448,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 19533
    },
    {
      "epoch": 4.650360054752127,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 19534
    },
    {
      "epoch": 4.650598107480807,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 19535
    },
    {
      "epoch": 4.6508361602094865,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 19536
    },
    {
      "epoch": 4.651074212938166,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 19537
    },
    {
      "epoch": 4.651312265666845,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 19538
    },
    {
      "epoch": 4.6515503183955245,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 19539
    },
    {
      "epoch": 4.651788371124204,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 19540
    },
    {
      "epoch": 4.652026423852884,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 19541
    },
    {
      "epoch": 4.652264476581562,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 19542
    },
    {
      "epoch": 4.652502529310242,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 19543
    },
    {
      "epoch": 4.652740582038922,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 19544
    },
    {
      "epoch": 4.652978634767601,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 19545
    },
    {
      "epoch": 4.653216687496281,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 19546
    },
    {
      "epoch": 4.6534547402249595,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 19547
    },
    {
      "epoch": 4.653692792953639,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 19548
    },
    {
      "epoch": 4.653930845682319,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 19549
    },
    {
      "epoch": 4.654168898410998,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 19550
    },
    {
      "epoch": 4.654406951139677,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 19551
    },
    {
      "epoch": 4.654645003868357,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 19552
    },
    {
      "epoch": 4.654883056597036,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 19553
    },
    {
      "epoch": 4.655121109325716,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 19554
    },
    {
      "epoch": 4.6553591620543955,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 19555
    },
    {
      "epoch": 4.655597214783074,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 19556
    },
    {
      "epoch": 4.655835267511754,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 19557
    },
    {
      "epoch": 4.656073320240433,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 19558
    },
    {
      "epoch": 4.656311372969113,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 19559
    },
    {
      "epoch": 4.656549425697792,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 19560
    },
    {
      "epoch": 4.656787478426471,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3712,
      "step": 19561
    },
    {
      "epoch": 4.657025531155151,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 19562
    },
    {
      "epoch": 4.6572635838838305,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 19563
    },
    {
      "epoch": 4.65750163661251,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 19564
    },
    {
      "epoch": 4.657739689341189,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 19565
    },
    {
      "epoch": 4.6579777420698685,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4803,
      "step": 19566
    },
    {
      "epoch": 4.658215794798548,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 19567
    },
    {
      "epoch": 4.658453847527227,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 19568
    },
    {
      "epoch": 4.658691900255906,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3666,
      "step": 19569
    },
    {
      "epoch": 4.658929952984586,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 19570
    },
    {
      "epoch": 4.659168005713266,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 19571
    },
    {
      "epoch": 4.659406058441945,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 19572
    },
    {
      "epoch": 4.659644111170624,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3025,
      "step": 19573
    },
    {
      "epoch": 4.6598821638993035,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 19574
    },
    {
      "epoch": 4.660120216627983,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3796,
      "step": 19575
    },
    {
      "epoch": 4.660358269356663,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 19576
    },
    {
      "epoch": 4.6605963220853415,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 19577
    },
    {
      "epoch": 4.660834374814021,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 19578
    },
    {
      "epoch": 4.661072427542701,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 19579
    },
    {
      "epoch": 4.66131048027138,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 19580
    },
    {
      "epoch": 4.66154853300006,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 19581
    },
    {
      "epoch": 4.661786585728739,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.4118,
      "step": 19582
    },
    {
      "epoch": 4.662024638457418,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 19583
    },
    {
      "epoch": 4.662262691186098,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 19584
    },
    {
      "epoch": 4.662500743914777,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8323,
      "step": 19585
    },
    {
      "epoch": 4.662738796643456,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 19586
    },
    {
      "epoch": 4.662976849372136,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 19587
    },
    {
      "epoch": 4.663214902100815,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 19588
    },
    {
      "epoch": 4.663452954829495,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 19589
    },
    {
      "epoch": 4.663691007558175,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 19590
    },
    {
      "epoch": 4.663929060286853,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 19591
    },
    {
      "epoch": 4.664167113015533,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 19592
    },
    {
      "epoch": 4.6644051657442125,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 19593
    },
    {
      "epoch": 4.664643218472892,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 19594
    },
    {
      "epoch": 4.664881271201571,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 19595
    },
    {
      "epoch": 4.66511932393025,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4123,
      "step": 19596
    },
    {
      "epoch": 4.66535737665893,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 19597
    },
    {
      "epoch": 4.66559542938761,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 19598
    },
    {
      "epoch": 4.665833482116289,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 19599
    },
    {
      "epoch": 4.666071534844968,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 19600
    },
    {
      "epoch": 4.666071534844968,
      "eval_runtime": 0.538,
      "eval_samples_per_second": 1.859,
      "eval_steps_per_second": 1.859,
      "step": 19600
    },
    {
      "epoch": 4.6663095875736476,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 19601
    },
    {
      "epoch": 4.666547640302327,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.426,
      "step": 19602
    },
    {
      "epoch": 4.666785693031006,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 19603
    },
    {
      "epoch": 4.6670237457596855,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 19604
    },
    {
      "epoch": 4.667261798488365,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 19605
    },
    {
      "epoch": 4.667499851217045,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 19606
    },
    {
      "epoch": 4.667737903945724,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 19607
    },
    {
      "epoch": 4.667975956674403,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 19608
    },
    {
      "epoch": 4.668214009403083,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2726,
      "step": 19609
    },
    {
      "epoch": 4.668452062131762,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 19610
    },
    {
      "epoch": 4.668690114860442,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.538,
      "step": 19611
    },
    {
      "epoch": 4.6689281675891205,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 19612
    },
    {
      "epoch": 4.6691662203178,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 19613
    },
    {
      "epoch": 4.66940427304648,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3758,
      "step": 19614
    },
    {
      "epoch": 4.669642325775159,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 19615
    },
    {
      "epoch": 4.669880378503839,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 19616
    },
    {
      "epoch": 4.670118431232518,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 19617
    },
    {
      "epoch": 4.670356483961197,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 19618
    },
    {
      "epoch": 4.670594536689877,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 19619
    },
    {
      "epoch": 4.6708325894185565,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 19620
    },
    {
      "epoch": 4.671070642147235,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 19621
    },
    {
      "epoch": 4.671308694875915,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 19622
    },
    {
      "epoch": 4.671546747604594,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 19623
    },
    {
      "epoch": 4.671784800333274,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 19624
    },
    {
      "epoch": 4.672022853061954,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 19625
    },
    {
      "epoch": 4.672260905790632,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 19626
    },
    {
      "epoch": 4.672498958519312,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4465,
      "step": 19627
    },
    {
      "epoch": 4.672737011247992,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6944,
      "step": 19628
    },
    {
      "epoch": 4.672975063976671,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 19629
    },
    {
      "epoch": 4.67321311670535,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 19630
    },
    {
      "epoch": 4.6734511694340295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 19631
    },
    {
      "epoch": 4.673689222162709,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 19632
    },
    {
      "epoch": 4.673927274891389,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 19633
    },
    {
      "epoch": 4.674165327620068,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 19634
    },
    {
      "epoch": 4.674403380348747,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 19635
    },
    {
      "epoch": 4.674641433077427,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 19636
    },
    {
      "epoch": 4.674879485806106,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 19637
    },
    {
      "epoch": 4.675117538534785,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3908,
      "step": 19638
    },
    {
      "epoch": 4.675355591263465,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 19639
    },
    {
      "epoch": 4.675593643992144,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 19640
    },
    {
      "epoch": 4.675831696720824,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 19641
    },
    {
      "epoch": 4.676069749449503,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 19642
    },
    {
      "epoch": 4.676307802178182,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 19643
    },
    {
      "epoch": 4.676545854906862,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 19644
    },
    {
      "epoch": 4.676783907635541,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 19645
    },
    {
      "epoch": 4.677021960364221,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.425,
      "step": 19646
    },
    {
      "epoch": 4.6772600130929,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 19647
    },
    {
      "epoch": 4.677498065821579,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 19648
    },
    {
      "epoch": 4.677736118550259,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 19649
    },
    {
      "epoch": 4.677974171278938,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 19650
    },
    {
      "epoch": 4.678212224007618,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2919,
      "step": 19651
    },
    {
      "epoch": 4.678450276736297,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 19652
    },
    {
      "epoch": 4.678688329464976,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 19653
    },
    {
      "epoch": 4.678926382193656,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 19654
    },
    {
      "epoch": 4.679164434922336,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 19655
    },
    {
      "epoch": 4.679402487651014,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 19656
    },
    {
      "epoch": 4.679640540379694,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 19657
    },
    {
      "epoch": 4.6798785931083735,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 19658
    },
    {
      "epoch": 4.680116645837053,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 19659
    },
    {
      "epoch": 4.680354698565733,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 19660
    },
    {
      "epoch": 4.680592751294411,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 19661
    },
    {
      "epoch": 4.680830804023091,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 19662
    },
    {
      "epoch": 4.681068856751771,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 19663
    },
    {
      "epoch": 4.68130690948045,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 19664
    },
    {
      "epoch": 4.681544962209129,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 19665
    },
    {
      "epoch": 4.681783014937809,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 19666
    },
    {
      "epoch": 4.682021067666488,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 19667
    },
    {
      "epoch": 4.682259120395168,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 19668
    },
    {
      "epoch": 4.682497173123847,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 19669
    },
    {
      "epoch": 4.682735225852526,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 19670
    },
    {
      "epoch": 4.682973278581206,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 19671
    },
    {
      "epoch": 4.683211331309885,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3383,
      "step": 19672
    },
    {
      "epoch": 4.683449384038565,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 19673
    },
    {
      "epoch": 4.683687436767244,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 19674
    },
    {
      "epoch": 4.683925489495923,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 19675
    },
    {
      "epoch": 4.684163542224603,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 19676
    },
    {
      "epoch": 4.6844015949532825,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 19677
    },
    {
      "epoch": 4.684639647681961,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 19678
    },
    {
      "epoch": 4.684877700410641,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 19679
    },
    {
      "epoch": 4.68511575313932,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 19680
    },
    {
      "epoch": 4.685353805868,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 19681
    },
    {
      "epoch": 4.685591858596679,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 19682
    },
    {
      "epoch": 4.685829911325358,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 19683
    },
    {
      "epoch": 4.686067964054038,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 19684
    },
    {
      "epoch": 4.6863060167827175,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 19685
    },
    {
      "epoch": 4.686544069511397,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3677,
      "step": 19686
    },
    {
      "epoch": 4.686782122240076,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 19687
    },
    {
      "epoch": 4.6870201749687554,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 19688
    },
    {
      "epoch": 4.687258227697435,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 19689
    },
    {
      "epoch": 4.687496280426115,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 19690
    },
    {
      "epoch": 4.687734333154793,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 19691
    },
    {
      "epoch": 4.687972385883473,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 19692
    },
    {
      "epoch": 4.688210438612153,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 19693
    },
    {
      "epoch": 4.688448491340832,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 19694
    },
    {
      "epoch": 4.688686544069512,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 19695
    },
    {
      "epoch": 4.6889245967981905,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 19696
    },
    {
      "epoch": 4.68916264952687,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 19697
    },
    {
      "epoch": 4.68940070225555,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 19698
    },
    {
      "epoch": 4.689638754984229,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 19699
    },
    {
      "epoch": 4.689876807712908,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 19700
    },
    {
      "epoch": 4.689876807712908,
      "eval_runtime": 0.5265,
      "eval_samples_per_second": 1.899,
      "eval_steps_per_second": 1.899,
      "step": 19700
    },
    {
      "epoch": 4.690114860441588,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 19701
    },
    {
      "epoch": 4.690352913170267,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 19702
    },
    {
      "epoch": 4.690590965898947,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 19703
    },
    {
      "epoch": 4.6908290186276265,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 19704
    },
    {
      "epoch": 4.691067071356305,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 19705
    },
    {
      "epoch": 4.691305124084985,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 19706
    },
    {
      "epoch": 4.691543176813664,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 19707
    },
    {
      "epoch": 4.691781229542344,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 19708
    },
    {
      "epoch": 4.692019282271023,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 19709
    },
    {
      "epoch": 4.692257334999702,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3827,
      "step": 19710
    },
    {
      "epoch": 4.692495387728382,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 19711
    },
    {
      "epoch": 4.6927334404570615,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 19712
    },
    {
      "epoch": 4.692971493185741,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 19713
    },
    {
      "epoch": 4.69320954591442,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 19714
    },
    {
      "epoch": 4.6934475986430995,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 19715
    },
    {
      "epoch": 4.693685651371779,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3599,
      "step": 19716
    },
    {
      "epoch": 4.693923704100458,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 19717
    },
    {
      "epoch": 4.694161756829137,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 19718
    },
    {
      "epoch": 4.694399809557817,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 19719
    },
    {
      "epoch": 4.694637862286497,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 19720
    },
    {
      "epoch": 4.694875915015176,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 19721
    },
    {
      "epoch": 4.695113967743855,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 19722
    },
    {
      "epoch": 4.6953520204725345,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3732,
      "step": 19723
    },
    {
      "epoch": 4.695590073201214,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 19724
    },
    {
      "epoch": 4.695828125929894,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 19725
    },
    {
      "epoch": 4.6960661786585725,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3865,
      "step": 19726
    },
    {
      "epoch": 4.696304231387252,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 19727
    },
    {
      "epoch": 4.696542284115932,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 19728
    },
    {
      "epoch": 4.696780336844611,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 19729
    },
    {
      "epoch": 4.697018389573291,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 19730
    },
    {
      "epoch": 4.69725644230197,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 19731
    },
    {
      "epoch": 4.697494495030649,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2435,
      "step": 19732
    },
    {
      "epoch": 4.697732547759329,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 19733
    },
    {
      "epoch": 4.697970600488008,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 19734
    },
    {
      "epoch": 4.698208653216687,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 19735
    },
    {
      "epoch": 4.698446705945367,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 19736
    },
    {
      "epoch": 4.698684758674046,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 19737
    },
    {
      "epoch": 4.698922811402726,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 19738
    },
    {
      "epoch": 4.6991608641314055,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 19739
    },
    {
      "epoch": 4.699398916860084,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 19740
    },
    {
      "epoch": 4.699636969588764,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 19741
    },
    {
      "epoch": 4.6998750223174435,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 19742
    },
    {
      "epoch": 4.700113075046123,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 19743
    },
    {
      "epoch": 4.700351127774802,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 19744
    },
    {
      "epoch": 4.700589180503481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 19745
    },
    {
      "epoch": 4.700827233232161,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.227,
      "step": 19746
    },
    {
      "epoch": 4.701065285960841,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 19747
    },
    {
      "epoch": 4.70130333868952,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 19748
    },
    {
      "epoch": 4.701541391418199,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7733,
      "step": 19749
    },
    {
      "epoch": 4.7017794441468785,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 19750
    },
    {
      "epoch": 4.702017496875558,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 19751
    },
    {
      "epoch": 4.702255549604237,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 19752
    },
    {
      "epoch": 4.7024936023329165,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 19753
    },
    {
      "epoch": 4.702731655061596,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 19754
    },
    {
      "epoch": 4.702969707790276,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 19755
    },
    {
      "epoch": 4.703207760518955,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 19756
    },
    {
      "epoch": 4.703445813247634,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 19757
    },
    {
      "epoch": 4.703683865976314,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 19758
    },
    {
      "epoch": 4.703921918704993,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 19759
    },
    {
      "epoch": 4.704159971433673,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 19760
    },
    {
      "epoch": 4.7043980241623515,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 19761
    },
    {
      "epoch": 4.704636076891031,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 19762
    },
    {
      "epoch": 4.704874129619711,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 19763
    },
    {
      "epoch": 4.70511218234839,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 19764
    },
    {
      "epoch": 4.70535023507707,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 19765
    },
    {
      "epoch": 4.705588287805749,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5227,
      "step": 19766
    },
    {
      "epoch": 4.705826340534428,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 19767
    },
    {
      "epoch": 4.706064393263108,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3479,
      "step": 19768
    },
    {
      "epoch": 4.7063024459917875,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 19769
    },
    {
      "epoch": 4.706540498720466,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 19770
    },
    {
      "epoch": 4.706778551449146,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 19771
    },
    {
      "epoch": 4.707016604177825,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 19772
    },
    {
      "epoch": 4.707254656906505,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 19773
    },
    {
      "epoch": 4.707492709635185,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4401,
      "step": 19774
    },
    {
      "epoch": 4.707730762363863,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 19775
    },
    {
      "epoch": 4.707968815092543,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 19776
    },
    {
      "epoch": 4.7082068678212226,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 19777
    },
    {
      "epoch": 4.708444920549902,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 19778
    },
    {
      "epoch": 4.708682973278581,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2921,
      "step": 19779
    },
    {
      "epoch": 4.7089210260072605,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 19780
    },
    {
      "epoch": 4.70915907873594,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 19781
    },
    {
      "epoch": 4.70939713146462,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 19782
    },
    {
      "epoch": 4.709635184193299,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 19783
    },
    {
      "epoch": 4.709873236921978,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 19784
    },
    {
      "epoch": 4.710111289650658,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 19785
    },
    {
      "epoch": 4.710349342379337,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 19786
    },
    {
      "epoch": 4.710587395108016,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 19787
    },
    {
      "epoch": 4.7108254478366955,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.332,
      "step": 19788
    },
    {
      "epoch": 4.711063500565375,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3725,
      "step": 19789
    },
    {
      "epoch": 4.711301553294055,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 19790
    },
    {
      "epoch": 4.711539606022734,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 19791
    },
    {
      "epoch": 4.711777658751413,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 19792
    },
    {
      "epoch": 4.712015711480093,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2487,
      "step": 19793
    },
    {
      "epoch": 4.712253764208772,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 19794
    },
    {
      "epoch": 4.712491816937452,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 19795
    },
    {
      "epoch": 4.712729869666131,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 19796
    },
    {
      "epoch": 4.71296792239481,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 19797
    },
    {
      "epoch": 4.71320597512349,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 19798
    },
    {
      "epoch": 4.713444027852169,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 19799
    },
    {
      "epoch": 4.713682080580849,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 19800
    },
    {
      "epoch": 4.713682080580849,
      "eval_runtime": 0.5442,
      "eval_samples_per_second": 1.838,
      "eval_steps_per_second": 1.838,
      "step": 19800
    },
    {
      "epoch": 4.713920133309528,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 19801
    },
    {
      "epoch": 4.714158186038207,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 19802
    },
    {
      "epoch": 4.714396238766887,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 19803
    },
    {
      "epoch": 4.714634291495567,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 19804
    },
    {
      "epoch": 4.714872344224245,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.672,
      "step": 19805
    },
    {
      "epoch": 4.715110396952925,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7887,
      "step": 19806
    },
    {
      "epoch": 4.7153484496816045,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 19807
    },
    {
      "epoch": 4.715586502410284,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7989,
      "step": 19808
    },
    {
      "epoch": 4.715824555138964,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3427,
      "step": 19809
    },
    {
      "epoch": 4.716062607867642,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3445,
      "step": 19810
    },
    {
      "epoch": 4.716300660596322,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 19811
    },
    {
      "epoch": 4.716538713325002,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 19812
    },
    {
      "epoch": 4.716776766053681,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 19813
    },
    {
      "epoch": 4.71701481878236,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 19814
    },
    {
      "epoch": 4.71725287151104,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 19815
    },
    {
      "epoch": 4.717490924239719,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 19816
    },
    {
      "epoch": 4.717728976968399,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 19817
    },
    {
      "epoch": 4.717967029697078,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 19818
    },
    {
      "epoch": 4.718205082425757,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3113,
      "step": 19819
    },
    {
      "epoch": 4.718443135154437,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 19820
    },
    {
      "epoch": 4.718681187883116,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 19821
    },
    {
      "epoch": 4.718919240611795,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 19822
    },
    {
      "epoch": 4.719157293340475,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 19823
    },
    {
      "epoch": 4.719395346069154,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 19824
    },
    {
      "epoch": 4.719633398797834,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 19825
    },
    {
      "epoch": 4.719871451526513,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 19826
    },
    {
      "epoch": 4.720109504255192,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 19827
    },
    {
      "epoch": 4.720347556983872,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 19828
    },
    {
      "epoch": 4.720585609712551,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 19829
    },
    {
      "epoch": 4.720823662441231,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 19830
    },
    {
      "epoch": 4.72106171516991,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 19831
    },
    {
      "epoch": 4.721299767898589,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 19832
    },
    {
      "epoch": 4.721537820627269,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 19833
    },
    {
      "epoch": 4.7217758733559485,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 19834
    },
    {
      "epoch": 4.722013926084628,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.4522,
      "step": 19835
    },
    {
      "epoch": 4.722251978813307,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 19836
    },
    {
      "epoch": 4.722490031541986,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 19837
    },
    {
      "epoch": 4.722728084270666,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8225,
      "step": 19838
    },
    {
      "epoch": 4.722966136999346,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 19839
    },
    {
      "epoch": 4.723204189728024,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 19840
    },
    {
      "epoch": 4.723442242456704,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1996,
      "step": 19841
    },
    {
      "epoch": 4.723680295185384,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 19842
    },
    {
      "epoch": 4.723918347914063,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 19843
    },
    {
      "epoch": 4.724156400642743,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 19844
    },
    {
      "epoch": 4.7243944533714215,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 19845
    },
    {
      "epoch": 4.724632506100101,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 19846
    },
    {
      "epoch": 4.724870558828781,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 19847
    },
    {
      "epoch": 4.72510861155746,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 19848
    },
    {
      "epoch": 4.725346664286139,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4972,
      "step": 19849
    },
    {
      "epoch": 4.725584717014819,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4687,
      "step": 19850
    },
    {
      "epoch": 4.725822769743498,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5643,
      "step": 19851
    },
    {
      "epoch": 4.726060822472178,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 19852
    },
    {
      "epoch": 4.7262988752008575,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 19853
    },
    {
      "epoch": 4.726536927929536,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 19854
    },
    {
      "epoch": 4.726774980658216,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 19855
    },
    {
      "epoch": 4.727013033386895,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 19856
    },
    {
      "epoch": 4.727251086115575,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 19857
    },
    {
      "epoch": 4.727489138844254,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0252,
      "step": 19858
    },
    {
      "epoch": 4.727727191572933,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 19859
    },
    {
      "epoch": 4.727965244301613,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 19860
    },
    {
      "epoch": 4.7282032970302925,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 19861
    },
    {
      "epoch": 4.728441349758971,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 19862
    },
    {
      "epoch": 4.728679402487651,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 19863
    },
    {
      "epoch": 4.7289174552163304,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 19864
    },
    {
      "epoch": 4.72915550794501,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8144,
      "step": 19865
    },
    {
      "epoch": 4.729393560673689,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 19866
    },
    {
      "epoch": 4.729631613402368,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3794,
      "step": 19867
    },
    {
      "epoch": 4.729869666131048,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 19868
    },
    {
      "epoch": 4.730107718859728,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 19869
    },
    {
      "epoch": 4.730345771588407,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 19870
    },
    {
      "epoch": 4.730583824317086,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 19871
    },
    {
      "epoch": 4.7308218770457655,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 19872
    },
    {
      "epoch": 4.731059929774445,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 19873
    },
    {
      "epoch": 4.731297982503125,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 19874
    },
    {
      "epoch": 4.731536035231803,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 19875
    },
    {
      "epoch": 4.731774087960483,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4347,
      "step": 19876
    },
    {
      "epoch": 4.732012140689163,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 19877
    },
    {
      "epoch": 4.732250193417842,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 19878
    },
    {
      "epoch": 4.732488246146522,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3831,
      "step": 19879
    },
    {
      "epoch": 4.732726298875201,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 19880
    },
    {
      "epoch": 4.73296435160388,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 19881
    },
    {
      "epoch": 4.73320240433256,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 19882
    },
    {
      "epoch": 4.733440457061239,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 19883
    },
    {
      "epoch": 4.733678509789918,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 19884
    },
    {
      "epoch": 4.733916562518598,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2428,
      "step": 19885
    },
    {
      "epoch": 4.734154615247277,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2912,
      "step": 19886
    },
    {
      "epoch": 4.734392667975957,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.349,
      "step": 19887
    },
    {
      "epoch": 4.7346307207046365,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 19888
    },
    {
      "epoch": 4.734868773433315,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 19889
    },
    {
      "epoch": 4.735106826161995,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 19890
    },
    {
      "epoch": 4.7353448788906745,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 19891
    },
    {
      "epoch": 4.735582931619354,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 19892
    },
    {
      "epoch": 4.735820984348033,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 19893
    },
    {
      "epoch": 4.736059037076712,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 19894
    },
    {
      "epoch": 4.736297089805392,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 19895
    },
    {
      "epoch": 4.736535142534072,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 19896
    },
    {
      "epoch": 4.73677319526275,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 19897
    },
    {
      "epoch": 4.73701124799143,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 19898
    },
    {
      "epoch": 4.7372493007201095,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4431,
      "step": 19899
    },
    {
      "epoch": 4.737487353448789,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 19900
    },
    {
      "epoch": 4.737487353448789,
      "eval_runtime": 0.517,
      "eval_samples_per_second": 1.934,
      "eval_steps_per_second": 1.934,
      "step": 19900
    },
    {
      "epoch": 4.737725406177468,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 19901
    },
    {
      "epoch": 4.7379634589061475,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4155,
      "step": 19902
    },
    {
      "epoch": 4.738201511634827,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 19903
    },
    {
      "epoch": 4.738439564363507,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 19904
    },
    {
      "epoch": 4.738677617092186,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.4577,
      "step": 19905
    },
    {
      "epoch": 4.738915669820865,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 19906
    },
    {
      "epoch": 4.739153722549545,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 19907
    },
    {
      "epoch": 4.739391775278224,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 19908
    },
    {
      "epoch": 4.739629828006904,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 19909
    },
    {
      "epoch": 4.7398678807355825,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2544,
      "step": 19910
    },
    {
      "epoch": 4.740105933464262,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3108,
      "step": 19911
    },
    {
      "epoch": 4.740343986192942,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.6888,
      "step": 19912
    },
    {
      "epoch": 4.740582038921621,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 19913
    },
    {
      "epoch": 4.740820091650301,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 19914
    },
    {
      "epoch": 4.74105814437898,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 19915
    },
    {
      "epoch": 4.741296197107659,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 19916
    },
    {
      "epoch": 4.741534249836339,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 19917
    },
    {
      "epoch": 4.7417723025650185,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 19918
    },
    {
      "epoch": 4.742010355293697,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 19919
    },
    {
      "epoch": 4.742248408022377,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 19920
    },
    {
      "epoch": 4.742486460751056,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 19921
    },
    {
      "epoch": 4.742724513479736,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2502,
      "step": 19922
    },
    {
      "epoch": 4.742962566208416,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3392,
      "step": 19923
    },
    {
      "epoch": 4.743200618937094,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2808,
      "step": 19924
    },
    {
      "epoch": 4.743438671665774,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 19925
    },
    {
      "epoch": 4.7436767243944535,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 19926
    },
    {
      "epoch": 4.743914777123133,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 19927
    },
    {
      "epoch": 4.744152829851812,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3615,
      "step": 19928
    },
    {
      "epoch": 4.7443908825804915,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 19929
    },
    {
      "epoch": 4.744628935309171,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 19930
    },
    {
      "epoch": 4.744866988037851,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 19931
    },
    {
      "epoch": 4.74510504076653,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 19932
    },
    {
      "epoch": 4.745343093495209,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 19933
    },
    {
      "epoch": 4.745581146223889,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 19934
    },
    {
      "epoch": 4.745819198952568,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 19935
    },
    {
      "epoch": 4.746057251681247,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 19936
    },
    {
      "epoch": 4.7462953044099265,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 19937
    },
    {
      "epoch": 4.746533357138606,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 19938
    },
    {
      "epoch": 4.746771409867286,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 19939
    },
    {
      "epoch": 4.747009462595965,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 19940
    },
    {
      "epoch": 4.747247515324644,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 19941
    },
    {
      "epoch": 4.747485568053324,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 19942
    },
    {
      "epoch": 4.747723620782003,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 19943
    },
    {
      "epoch": 4.747961673510683,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 19944
    },
    {
      "epoch": 4.748199726239362,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 19945
    },
    {
      "epoch": 4.748437778968041,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 19946
    },
    {
      "epoch": 4.748675831696721,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4724,
      "step": 19947
    },
    {
      "epoch": 4.7489138844254,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 19948
    },
    {
      "epoch": 4.74915193715408,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4374,
      "step": 19949
    },
    {
      "epoch": 4.749389989882759,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4321,
      "step": 19950
    },
    {
      "epoch": 4.749628042611438,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 19951
    },
    {
      "epoch": 4.749866095340118,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9831,
      "step": 19952
    },
    {
      "epoch": 4.7501041480687975,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 19953
    },
    {
      "epoch": 4.750342200797476,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 19954
    },
    {
      "epoch": 4.750580253526156,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 19955
    },
    {
      "epoch": 4.7508183062548355,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 19956
    },
    {
      "epoch": 4.751056358983515,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 19957
    },
    {
      "epoch": 4.751294411712195,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.6548,
      "step": 19958
    },
    {
      "epoch": 4.751532464440873,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 19959
    },
    {
      "epoch": 4.751770517169553,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 19960
    },
    {
      "epoch": 4.752008569898233,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 19961
    },
    {
      "epoch": 4.752246622626912,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 19962
    },
    {
      "epoch": 4.752484675355591,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 19963
    },
    {
      "epoch": 4.7527227280842705,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 19964
    },
    {
      "epoch": 4.75296078081295,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 19965
    },
    {
      "epoch": 4.75319883354163,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 19966
    },
    {
      "epoch": 4.753436886270309,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 19967
    },
    {
      "epoch": 4.753674938998988,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 19968
    },
    {
      "epoch": 4.753912991727668,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 19969
    },
    {
      "epoch": 4.754151044456347,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 19970
    },
    {
      "epoch": 4.754389097185026,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 19971
    },
    {
      "epoch": 4.754627149913706,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 19972
    },
    {
      "epoch": 4.754865202642385,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 19973
    },
    {
      "epoch": 4.755103255371065,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 19974
    },
    {
      "epoch": 4.755341308099744,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 19975
    },
    {
      "epoch": 4.755579360828423,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 19976
    },
    {
      "epoch": 4.755817413557103,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 19977
    },
    {
      "epoch": 4.756055466285782,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2357,
      "step": 19978
    },
    {
      "epoch": 4.756293519014462,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 19979
    },
    {
      "epoch": 4.756531571743141,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.91,
      "step": 19980
    },
    {
      "epoch": 4.75676962447182,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 19981
    },
    {
      "epoch": 4.7570076772005,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 19982
    },
    {
      "epoch": 4.7572457299291795,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 19983
    },
    {
      "epoch": 4.757483782657859,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 19984
    },
    {
      "epoch": 4.757721835386538,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 19985
    },
    {
      "epoch": 4.757959888115217,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 19986
    },
    {
      "epoch": 4.758197940843897,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 19987
    },
    {
      "epoch": 4.758435993572577,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 19988
    },
    {
      "epoch": 4.758674046301255,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 19989
    },
    {
      "epoch": 4.758912099029935,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 19990
    },
    {
      "epoch": 4.7591501517586146,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.261,
      "step": 19991
    },
    {
      "epoch": 4.759388204487294,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 19992
    },
    {
      "epoch": 4.759626257215974,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.5469,
      "step": 19993
    },
    {
      "epoch": 4.7598643099446525,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3532,
      "step": 19994
    },
    {
      "epoch": 4.760102362673332,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 19995
    },
    {
      "epoch": 4.760340415402012,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 19996
    },
    {
      "epoch": 4.760578468130691,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 19997
    },
    {
      "epoch": 4.76081652085937,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 19998
    },
    {
      "epoch": 4.76105457358805,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 19999
    },
    {
      "epoch": 4.761292626316729,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 20000
    },
    {
      "epoch": 4.761292626316729,
      "eval_runtime": 0.5388,
      "eval_samples_per_second": 1.856,
      "eval_steps_per_second": 1.856,
      "step": 20000
    },
    {
      "epoch": 4.761530679045409,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2395,
      "step": 20001
    },
    {
      "epoch": 4.761768731774088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 20002
    },
    {
      "epoch": 4.762006784502767,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 20003
    },
    {
      "epoch": 4.762244837231447,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 20004
    },
    {
      "epoch": 4.762482889960126,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 20005
    },
    {
      "epoch": 4.762720942688805,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 20006
    },
    {
      "epoch": 4.762958995417485,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 20007
    },
    {
      "epoch": 4.763197048146164,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.4853,
      "step": 20008
    },
    {
      "epoch": 4.763435100874844,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 20009
    },
    {
      "epoch": 4.7636731536035235,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 20010
    },
    {
      "epoch": 4.763911206332202,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 20011
    },
    {
      "epoch": 4.764149259060882,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 20012
    },
    {
      "epoch": 4.764387311789561,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 20013
    },
    {
      "epoch": 4.764625364518241,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 20014
    },
    {
      "epoch": 4.76486341724692,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 20015
    },
    {
      "epoch": 4.765101469975599,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 20016
    },
    {
      "epoch": 4.765339522704279,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 20017
    },
    {
      "epoch": 4.765577575432959,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 20018
    },
    {
      "epoch": 4.765815628161638,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 20019
    },
    {
      "epoch": 4.766053680890317,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3734,
      "step": 20020
    },
    {
      "epoch": 4.7662917336189965,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2474,
      "step": 20021
    },
    {
      "epoch": 4.766529786347676,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 20022
    },
    {
      "epoch": 4.766767839076356,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 20023
    },
    {
      "epoch": 4.767005891805034,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 20024
    },
    {
      "epoch": 4.767243944533714,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 20025
    },
    {
      "epoch": 4.767481997262394,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 20026
    },
    {
      "epoch": 4.767720049991073,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 20027
    },
    {
      "epoch": 4.767958102719753,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 20028
    },
    {
      "epoch": 4.768196155448432,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 20029
    },
    {
      "epoch": 4.768434208177111,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 20030
    },
    {
      "epoch": 4.768672260905791,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 20031
    },
    {
      "epoch": 4.76891031363447,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 20032
    },
    {
      "epoch": 4.769148366363149,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 20033
    },
    {
      "epoch": 4.769386419091829,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 20034
    },
    {
      "epoch": 4.769624471820508,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3611,
      "step": 20035
    },
    {
      "epoch": 4.769862524549188,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 20036
    },
    {
      "epoch": 4.7701005772778675,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 20037
    },
    {
      "epoch": 4.770338630006546,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 20038
    },
    {
      "epoch": 4.770576682735226,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 20039
    },
    {
      "epoch": 4.770814735463905,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 20040
    },
    {
      "epoch": 4.771052788192584,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 20041
    },
    {
      "epoch": 4.771290840921264,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 20042
    },
    {
      "epoch": 4.771528893649943,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3963,
      "step": 20043
    },
    {
      "epoch": 4.771766946378623,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 20044
    },
    {
      "epoch": 4.772004999107303,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 20045
    },
    {
      "epoch": 4.772243051835981,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 20046
    },
    {
      "epoch": 4.772481104564661,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2082,
      "step": 20047
    },
    {
      "epoch": 4.7727191572933405,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 20048
    },
    {
      "epoch": 4.77295721002202,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 20049
    },
    {
      "epoch": 4.773195262750699,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 20050
    },
    {
      "epoch": 4.773433315479378,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 20051
    },
    {
      "epoch": 4.773671368208058,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3407,
      "step": 20052
    },
    {
      "epoch": 4.773909420936738,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 20053
    },
    {
      "epoch": 4.774147473665417,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 20054
    },
    {
      "epoch": 4.774385526394096,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 20055
    },
    {
      "epoch": 4.774623579122776,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 20056
    },
    {
      "epoch": 4.774861631851455,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 20057
    },
    {
      "epoch": 4.775099684580135,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 20058
    },
    {
      "epoch": 4.7753377373088135,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 20059
    },
    {
      "epoch": 4.775575790037493,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 20060
    },
    {
      "epoch": 4.775813842766173,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 20061
    },
    {
      "epoch": 4.776051895494852,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 20062
    },
    {
      "epoch": 4.776289948223532,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 20063
    },
    {
      "epoch": 4.776528000952211,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3321,
      "step": 20064
    },
    {
      "epoch": 4.77676605368089,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.289,
      "step": 20065
    },
    {
      "epoch": 4.77700410640957,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 20066
    },
    {
      "epoch": 4.7772421591382495,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 20067
    },
    {
      "epoch": 4.777480211866928,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 20068
    },
    {
      "epoch": 4.777718264595608,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.6417,
      "step": 20069
    },
    {
      "epoch": 4.777956317324287,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4634,
      "step": 20070
    },
    {
      "epoch": 4.778194370052967,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 20071
    },
    {
      "epoch": 4.778432422781647,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 20072
    },
    {
      "epoch": 4.778670475510325,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 20073
    },
    {
      "epoch": 4.778908528239005,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 20074
    },
    {
      "epoch": 4.7791465809676845,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 20075
    },
    {
      "epoch": 4.779384633696364,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.5496,
      "step": 20076
    },
    {
      "epoch": 4.779622686425043,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5358,
      "step": 20077
    },
    {
      "epoch": 4.7798607391537224,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 20078
    },
    {
      "epoch": 4.780098791882402,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 20079
    },
    {
      "epoch": 4.780336844611082,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 20080
    },
    {
      "epoch": 4.78057489733976,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 20081
    },
    {
      "epoch": 4.78081295006844,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 20082
    },
    {
      "epoch": 4.78105100279712,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 20083
    },
    {
      "epoch": 4.781289055525799,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 20084
    },
    {
      "epoch": 4.781527108254478,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 20085
    },
    {
      "epoch": 4.7817651609831575,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3483,
      "step": 20086
    },
    {
      "epoch": 4.782003213711837,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 20087
    },
    {
      "epoch": 4.782241266440517,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 20088
    },
    {
      "epoch": 4.782479319169196,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 20089
    },
    {
      "epoch": 4.782717371897875,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 20090
    },
    {
      "epoch": 4.782955424626555,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 20091
    },
    {
      "epoch": 4.783193477355234,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4903,
      "step": 20092
    },
    {
      "epoch": 4.783431530083914,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 20093
    },
    {
      "epoch": 4.783669582812593,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 20094
    },
    {
      "epoch": 4.783907635541272,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 20095
    },
    {
      "epoch": 4.784145688269952,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 20096
    },
    {
      "epoch": 4.784383740998631,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 20097
    },
    {
      "epoch": 4.784621793727311,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 20098
    },
    {
      "epoch": 4.78485984645599,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 20099
    },
    {
      "epoch": 4.785097899184669,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 20100
    },
    {
      "epoch": 4.785097899184669,
      "eval_runtime": 0.5354,
      "eval_samples_per_second": 1.868,
      "eval_steps_per_second": 1.868,
      "step": 20100
    },
    {
      "epoch": 4.785335951913349,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.295,
      "step": 20101
    },
    {
      "epoch": 4.7855740046420285,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 20102
    },
    {
      "epoch": 4.785812057370707,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2511,
      "step": 20103
    },
    {
      "epoch": 4.786050110099387,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 20104
    },
    {
      "epoch": 4.7862881628280665,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4992,
      "step": 20105
    },
    {
      "epoch": 4.786526215556746,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 20106
    },
    {
      "epoch": 4.786764268285426,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 20107
    },
    {
      "epoch": 4.787002321014104,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 20108
    },
    {
      "epoch": 4.787240373742784,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 20109
    },
    {
      "epoch": 4.787478426471464,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2693,
      "step": 20110
    },
    {
      "epoch": 4.787716479200143,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 20111
    },
    {
      "epoch": 4.787954531928822,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 20112
    },
    {
      "epoch": 4.7881925846575015,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 20113
    },
    {
      "epoch": 4.788430637386181,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 20114
    },
    {
      "epoch": 4.788668690114861,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 20115
    },
    {
      "epoch": 4.78890674284354,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 20116
    },
    {
      "epoch": 4.789144795572219,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 20117
    },
    {
      "epoch": 4.789382848300899,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3282,
      "step": 20118
    },
    {
      "epoch": 4.789620901029578,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 20119
    },
    {
      "epoch": 4.789858953758257,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4508,
      "step": 20120
    },
    {
      "epoch": 4.790097006486937,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 20121
    },
    {
      "epoch": 4.790335059215616,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 20122
    },
    {
      "epoch": 4.790573111944296,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 20123
    },
    {
      "epoch": 4.790811164672975,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 20124
    },
    {
      "epoch": 4.791049217401654,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 20125
    },
    {
      "epoch": 4.791287270130334,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 20126
    },
    {
      "epoch": 4.791525322859013,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3824,
      "step": 20127
    },
    {
      "epoch": 4.791763375587693,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 20128
    },
    {
      "epoch": 4.792001428316372,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 20129
    },
    {
      "epoch": 4.792239481045051,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 20130
    },
    {
      "epoch": 4.792477533773731,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 20131
    },
    {
      "epoch": 4.7927155865024105,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3227,
      "step": 20132
    },
    {
      "epoch": 4.79295363923109,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 20133
    },
    {
      "epoch": 4.793191691959769,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 20134
    },
    {
      "epoch": 4.793429744688448,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 20135
    },
    {
      "epoch": 4.793667797417128,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 20136
    },
    {
      "epoch": 4.793905850145808,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2166,
      "step": 20137
    },
    {
      "epoch": 4.794143902874486,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 20138
    },
    {
      "epoch": 4.794381955603166,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7879,
      "step": 20139
    },
    {
      "epoch": 4.7946200083318455,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 20140
    },
    {
      "epoch": 4.794858061060525,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 20141
    },
    {
      "epoch": 4.795096113789205,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 20142
    },
    {
      "epoch": 4.7953341665178835,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 20143
    },
    {
      "epoch": 4.795572219246563,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 20144
    },
    {
      "epoch": 4.795810271975243,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 20145
    },
    {
      "epoch": 4.796048324703922,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 20146
    },
    {
      "epoch": 4.796286377432601,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 20147
    },
    {
      "epoch": 4.796524430161281,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7677,
      "step": 20148
    },
    {
      "epoch": 4.79676248288996,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 20149
    },
    {
      "epoch": 4.79700053561864,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 20150
    },
    {
      "epoch": 4.797238588347319,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 20151
    },
    {
      "epoch": 4.797476641075998,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 20152
    },
    {
      "epoch": 4.797714693804678,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 20153
    },
    {
      "epoch": 4.797952746533357,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 20154
    },
    {
      "epoch": 4.798190799262036,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4564,
      "step": 20155
    },
    {
      "epoch": 4.798428851990716,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 20156
    },
    {
      "epoch": 4.798666904719395,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5011,
      "step": 20157
    },
    {
      "epoch": 4.798904957448075,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 20158
    },
    {
      "epoch": 4.7991430101767545,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 20159
    },
    {
      "epoch": 4.799381062905433,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 20160
    },
    {
      "epoch": 4.799619115634113,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 20161
    },
    {
      "epoch": 4.799857168362792,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 20162
    },
    {
      "epoch": 4.800095221091472,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 20163
    },
    {
      "epoch": 4.800333273820151,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 20164
    },
    {
      "epoch": 4.80057132654883,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 20165
    },
    {
      "epoch": 4.80080937927751,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 20166
    },
    {
      "epoch": 4.8010474320061896,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.079,
      "step": 20167
    },
    {
      "epoch": 4.801285484734869,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 20168
    },
    {
      "epoch": 4.801523537463548,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 20169
    },
    {
      "epoch": 4.8017615901922275,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 20170
    },
    {
      "epoch": 4.801999642920907,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 20171
    },
    {
      "epoch": 4.802237695649587,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 20172
    },
    {
      "epoch": 4.802475748378265,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 20173
    },
    {
      "epoch": 4.802713801106945,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 20174
    },
    {
      "epoch": 4.802951853835625,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 20175
    },
    {
      "epoch": 4.803189906564304,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 20176
    },
    {
      "epoch": 4.803427959292984,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 20177
    },
    {
      "epoch": 4.8036660120216625,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3137,
      "step": 20178
    },
    {
      "epoch": 4.803904064750342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 20179
    },
    {
      "epoch": 4.804142117479022,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 20180
    },
    {
      "epoch": 4.804380170207701,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 20181
    },
    {
      "epoch": 4.80461822293638,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 20182
    },
    {
      "epoch": 4.80485627566506,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3166,
      "step": 20183
    },
    {
      "epoch": 4.805094328393739,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 20184
    },
    {
      "epoch": 4.805332381122419,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 20185
    },
    {
      "epoch": 4.8055704338510985,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 20186
    },
    {
      "epoch": 4.805808486579777,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 20187
    },
    {
      "epoch": 4.806046539308457,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 20188
    },
    {
      "epoch": 4.806284592037136,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3752,
      "step": 20189
    },
    {
      "epoch": 4.806522644765815,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 20190
    },
    {
      "epoch": 4.806760697494495,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 20191
    },
    {
      "epoch": 4.806998750223174,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.273,
      "step": 20192
    },
    {
      "epoch": 4.807236802951854,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 20193
    },
    {
      "epoch": 4.807474855680534,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 20194
    },
    {
      "epoch": 4.807712908409212,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 20195
    },
    {
      "epoch": 4.807950961137892,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 20196
    },
    {
      "epoch": 4.8081890138665715,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 20197
    },
    {
      "epoch": 4.808427066595251,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 20198
    },
    {
      "epoch": 4.80866511932393,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 20199
    },
    {
      "epoch": 4.808903172052609,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 20200
    },
    {
      "epoch": 4.808903172052609,
      "eval_runtime": 0.5284,
      "eval_samples_per_second": 1.892,
      "eval_steps_per_second": 1.892,
      "step": 20200
    },
    {
      "epoch": 4.809141224781289,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 20201
    },
    {
      "epoch": 4.809379277509969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 20202
    },
    {
      "epoch": 4.809617330238648,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 20203
    },
    {
      "epoch": 4.809855382967327,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 20204
    },
    {
      "epoch": 4.810093435696007,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 20205
    },
    {
      "epoch": 4.810331488424686,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 20206
    },
    {
      "epoch": 4.810569541153366,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 20207
    },
    {
      "epoch": 4.8108075938820445,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 20208
    },
    {
      "epoch": 4.811045646610724,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 20209
    },
    {
      "epoch": 4.811283699339404,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 20210
    },
    {
      "epoch": 4.811521752068083,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 20211
    },
    {
      "epoch": 4.811759804796763,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 20212
    },
    {
      "epoch": 4.811997857525442,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 20213
    },
    {
      "epoch": 4.812235910254121,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 20214
    },
    {
      "epoch": 4.812473962982801,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 20215
    },
    {
      "epoch": 4.81271201571148,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 20216
    },
    {
      "epoch": 4.812950068440159,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 20217
    },
    {
      "epoch": 4.813188121168839,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 20218
    },
    {
      "epoch": 4.813426173897518,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 20219
    },
    {
      "epoch": 4.813664226626198,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 20220
    },
    {
      "epoch": 4.813902279354878,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.953,
      "step": 20221
    },
    {
      "epoch": 4.814140332083556,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 20222
    },
    {
      "epoch": 4.814378384812236,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 20223
    },
    {
      "epoch": 4.8146164375409155,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4223,
      "step": 20224
    },
    {
      "epoch": 4.814854490269594,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 20225
    },
    {
      "epoch": 4.815092542998274,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 20226
    },
    {
      "epoch": 4.815330595726953,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 20227
    },
    {
      "epoch": 4.815568648455633,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 20228
    },
    {
      "epoch": 4.815806701184313,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 20229
    },
    {
      "epoch": 4.816044753912991,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 20230
    },
    {
      "epoch": 4.816282806641671,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 20231
    },
    {
      "epoch": 4.816520859370351,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 20232
    },
    {
      "epoch": 4.81675891209903,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 20233
    },
    {
      "epoch": 4.816996964827709,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 20234
    },
    {
      "epoch": 4.8172350175563885,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 20235
    },
    {
      "epoch": 4.817473070285068,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 20236
    },
    {
      "epoch": 4.817711123013748,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 20237
    },
    {
      "epoch": 4.817949175742427,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 20238
    },
    {
      "epoch": 4.818187228471106,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3421,
      "step": 20239
    },
    {
      "epoch": 4.818425281199786,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8165,
      "step": 20240
    },
    {
      "epoch": 4.818663333928465,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 20241
    },
    {
      "epoch": 4.818901386657145,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 20242
    },
    {
      "epoch": 4.819139439385824,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3978,
      "step": 20243
    },
    {
      "epoch": 4.819377492114503,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 20244
    },
    {
      "epoch": 4.819615544843183,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 20245
    },
    {
      "epoch": 4.819853597571862,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 20246
    },
    {
      "epoch": 4.820091650300542,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 20247
    },
    {
      "epoch": 4.820329703029221,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 20248
    },
    {
      "epoch": 4.8205677557579,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 20249
    },
    {
      "epoch": 4.82080580848658,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 20250
    },
    {
      "epoch": 4.8210438612152595,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 20251
    },
    {
      "epoch": 4.821281913943938,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 20252
    },
    {
      "epoch": 4.821519966672618,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 20253
    },
    {
      "epoch": 4.8217580194012974,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.3653,
      "step": 20254
    },
    {
      "epoch": 4.821996072129977,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 20255
    },
    {
      "epoch": 4.822234124858657,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 20256
    },
    {
      "epoch": 4.822472177587335,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 20257
    },
    {
      "epoch": 4.822710230316015,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 20258
    },
    {
      "epoch": 4.822948283044695,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 20259
    },
    {
      "epoch": 4.823186335773374,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 20260
    },
    {
      "epoch": 4.823424388502053,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 20261
    },
    {
      "epoch": 4.8236624412307325,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 20262
    },
    {
      "epoch": 4.823900493959412,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 20263
    },
    {
      "epoch": 4.824138546688092,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 20264
    },
    {
      "epoch": 4.82437659941677,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4402,
      "step": 20265
    },
    {
      "epoch": 4.82461465214545,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 20266
    },
    {
      "epoch": 4.82485270487413,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 20267
    },
    {
      "epoch": 4.825090757602809,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 20268
    },
    {
      "epoch": 4.825328810331488,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 20269
    },
    {
      "epoch": 4.825566863060168,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 20270
    },
    {
      "epoch": 4.825804915788847,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 20271
    },
    {
      "epoch": 4.826042968517527,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 20272
    },
    {
      "epoch": 4.826281021246206,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 20273
    },
    {
      "epoch": 4.826519073974885,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 20274
    },
    {
      "epoch": 4.826757126703565,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.367,
      "step": 20275
    },
    {
      "epoch": 4.826995179432244,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 20276
    },
    {
      "epoch": 4.827233232160924,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 20277
    },
    {
      "epoch": 4.827471284889603,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 20278
    },
    {
      "epoch": 4.827709337618282,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 20279
    },
    {
      "epoch": 4.827947390346962,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 20280
    },
    {
      "epoch": 4.8281854430756415,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 20281
    },
    {
      "epoch": 4.828423495804321,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 20282
    },
    {
      "epoch": 4.828661548533,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 20283
    },
    {
      "epoch": 4.828899601261679,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 20284
    },
    {
      "epoch": 4.829137653990359,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 20285
    },
    {
      "epoch": 4.829375706719039,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 20286
    },
    {
      "epoch": 4.829613759447717,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 20287
    },
    {
      "epoch": 4.829851812176397,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3899,
      "step": 20288
    },
    {
      "epoch": 4.8300898649050765,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 20289
    },
    {
      "epoch": 4.830327917633756,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 20290
    },
    {
      "epoch": 4.830565970362436,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 20291
    },
    {
      "epoch": 4.8308040230911145,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 20292
    },
    {
      "epoch": 4.831042075819794,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 20293
    },
    {
      "epoch": 4.831280128548474,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 20294
    },
    {
      "epoch": 4.831518181277153,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 20295
    },
    {
      "epoch": 4.831756234005832,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 20296
    },
    {
      "epoch": 4.831994286734512,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3059,
      "step": 20297
    },
    {
      "epoch": 4.832232339463191,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 20298
    },
    {
      "epoch": 4.832470392191871,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 20299
    },
    {
      "epoch": 4.8327084449205495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 20300
    },
    {
      "epoch": 4.8327084449205495,
      "eval_runtime": 0.5315,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 20300
    },
    {
      "epoch": 4.832946497649229,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 20301
    },
    {
      "epoch": 4.833184550377909,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8136,
      "step": 20302
    },
    {
      "epoch": 4.833422603106588,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 20303
    },
    {
      "epoch": 4.833660655835267,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.5845,
      "step": 20304
    },
    {
      "epoch": 4.833898708563947,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 20305
    },
    {
      "epoch": 4.834136761292626,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 20306
    },
    {
      "epoch": 4.834374814021306,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 20307
    },
    {
      "epoch": 4.8346128667499855,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 20308
    },
    {
      "epoch": 4.834850919478664,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 20309
    },
    {
      "epoch": 4.835088972207344,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 20310
    },
    {
      "epoch": 4.835327024936023,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 20311
    },
    {
      "epoch": 4.835565077664703,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 20312
    },
    {
      "epoch": 4.835803130393382,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 20313
    },
    {
      "epoch": 4.836041183122061,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.299,
      "step": 20314
    },
    {
      "epoch": 4.836279235850741,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 20315
    },
    {
      "epoch": 4.8365172885794205,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 20316
    },
    {
      "epoch": 4.8367553413081,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2705,
      "step": 20317
    },
    {
      "epoch": 4.836993394036779,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3963,
      "step": 20318
    },
    {
      "epoch": 4.8372314467654585,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 20319
    },
    {
      "epoch": 4.837469499494138,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 20320
    },
    {
      "epoch": 4.837707552222818,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 20321
    },
    {
      "epoch": 4.837945604951496,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 20322
    },
    {
      "epoch": 4.838183657680176,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 20323
    },
    {
      "epoch": 4.838421710408856,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 20324
    },
    {
      "epoch": 4.838659763137535,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 20325
    },
    {
      "epoch": 4.838897815866215,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 20326
    },
    {
      "epoch": 4.8391358685948935,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2654,
      "step": 20327
    },
    {
      "epoch": 4.839373921323573,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 20328
    },
    {
      "epoch": 4.839611974052253,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8615,
      "step": 20329
    },
    {
      "epoch": 4.839850026780932,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 20330
    },
    {
      "epoch": 4.840088079509611,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 20331
    },
    {
      "epoch": 4.840326132238291,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 20332
    },
    {
      "epoch": 4.84056418496697,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 20333
    },
    {
      "epoch": 4.84080223769565,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 20334
    },
    {
      "epoch": 4.8410402904243295,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3053,
      "step": 20335
    },
    {
      "epoch": 4.841278343153008,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 20336
    },
    {
      "epoch": 4.841516395881688,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 20337
    },
    {
      "epoch": 4.841754448610367,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4785,
      "step": 20338
    },
    {
      "epoch": 4.841992501339046,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 20339
    },
    {
      "epoch": 4.842230554067726,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 20340
    },
    {
      "epoch": 4.842468606796405,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 20341
    },
    {
      "epoch": 4.842706659525085,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 20342
    },
    {
      "epoch": 4.8429447122537645,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 20343
    },
    {
      "epoch": 4.843182764982443,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 20344
    },
    {
      "epoch": 4.843420817711123,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 20345
    },
    {
      "epoch": 4.8436588704398025,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 20346
    },
    {
      "epoch": 4.843896923168482,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 20347
    },
    {
      "epoch": 4.844134975897161,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 20348
    },
    {
      "epoch": 4.84437302862584,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 20349
    },
    {
      "epoch": 4.84461108135452,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 20350
    },
    {
      "epoch": 4.8448491340832,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 20351
    },
    {
      "epoch": 4.845087186811879,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 20352
    },
    {
      "epoch": 4.845325239540558,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 20353
    },
    {
      "epoch": 4.8455632922692375,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 20354
    },
    {
      "epoch": 4.845801344997917,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 20355
    },
    {
      "epoch": 4.846039397726597,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 20356
    },
    {
      "epoch": 4.8462774504552755,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 20357
    },
    {
      "epoch": 4.846515503183955,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 20358
    },
    {
      "epoch": 4.846753555912635,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 20359
    },
    {
      "epoch": 4.846991608641314,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 20360
    },
    {
      "epoch": 4.847229661369994,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 20361
    },
    {
      "epoch": 4.847467714098673,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 20362
    },
    {
      "epoch": 4.847705766827352,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 20363
    },
    {
      "epoch": 4.847943819556032,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 20364
    },
    {
      "epoch": 4.848181872284711,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 20365
    },
    {
      "epoch": 4.84841992501339,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 20366
    },
    {
      "epoch": 4.84865797774207,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 20367
    },
    {
      "epoch": 4.848896030470749,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 20368
    },
    {
      "epoch": 4.849134083199429,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 20369
    },
    {
      "epoch": 4.849372135928109,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 20370
    },
    {
      "epoch": 4.849610188656787,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 20371
    },
    {
      "epoch": 4.849848241385467,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 20372
    },
    {
      "epoch": 4.8500862941141465,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 20373
    },
    {
      "epoch": 4.850324346842825,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 20374
    },
    {
      "epoch": 4.850562399571505,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 20375
    },
    {
      "epoch": 4.850800452300184,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 20376
    },
    {
      "epoch": 4.851038505028864,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3658,
      "step": 20377
    },
    {
      "epoch": 4.851276557757544,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4199,
      "step": 20378
    },
    {
      "epoch": 4.851514610486222,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4526,
      "step": 20379
    },
    {
      "epoch": 4.851752663214902,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.4012,
      "step": 20380
    },
    {
      "epoch": 4.8519907159435816,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 20381
    },
    {
      "epoch": 4.852228768672261,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 20382
    },
    {
      "epoch": 4.85246682140094,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 20383
    },
    {
      "epoch": 4.8527048741296195,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 20384
    },
    {
      "epoch": 4.852942926858299,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 20385
    },
    {
      "epoch": 4.853180979586979,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 20386
    },
    {
      "epoch": 4.853419032315658,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 20387
    },
    {
      "epoch": 4.853657085044337,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 20388
    },
    {
      "epoch": 4.853895137773017,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 20389
    },
    {
      "epoch": 4.854133190501696,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 20390
    },
    {
      "epoch": 4.854371243230376,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 20391
    },
    {
      "epoch": 4.8546092959590545,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 20392
    },
    {
      "epoch": 4.854847348687734,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.5187,
      "step": 20393
    },
    {
      "epoch": 4.855085401416414,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 20394
    },
    {
      "epoch": 4.855323454145093,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 20395
    },
    {
      "epoch": 4.855561506873773,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 20396
    },
    {
      "epoch": 4.855799559602452,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4529,
      "step": 20397
    },
    {
      "epoch": 4.856037612331131,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 20398
    },
    {
      "epoch": 4.856275665059811,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 20399
    },
    {
      "epoch": 4.8565137177884905,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3072,
      "step": 20400
    },
    {
      "epoch": 4.8565137177884905,
      "eval_runtime": 0.5138,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 20400
    },
    {
      "epoch": 4.856751770517169,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 20401
    },
    {
      "epoch": 4.856989823245849,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 20402
    },
    {
      "epoch": 4.857227875974528,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 20403
    },
    {
      "epoch": 4.857465928703208,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 20404
    },
    {
      "epoch": 4.857703981431888,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 20405
    },
    {
      "epoch": 4.857942034160566,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 20406
    },
    {
      "epoch": 4.858180086889246,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 20407
    },
    {
      "epoch": 4.858418139617926,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 20408
    },
    {
      "epoch": 4.858656192346604,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 20409
    },
    {
      "epoch": 4.858894245075284,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 20410
    },
    {
      "epoch": 4.8591322978039635,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 20411
    },
    {
      "epoch": 4.859370350532643,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 20412
    },
    {
      "epoch": 4.859608403261323,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 20413
    },
    {
      "epoch": 4.859846455990001,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 20414
    },
    {
      "epoch": 4.860084508718681,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 20415
    },
    {
      "epoch": 4.860322561447361,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 20416
    },
    {
      "epoch": 4.86056061417604,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 20417
    },
    {
      "epoch": 4.860798666904719,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 20418
    },
    {
      "epoch": 4.861036719633399,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 20419
    },
    {
      "epoch": 4.861274772362078,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 20420
    },
    {
      "epoch": 4.861512825090758,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2852,
      "step": 20421
    },
    {
      "epoch": 4.861750877819437,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 20422
    },
    {
      "epoch": 4.861988930548116,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 20423
    },
    {
      "epoch": 4.862226983276796,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 20424
    },
    {
      "epoch": 4.862465036005475,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 20425
    },
    {
      "epoch": 4.862703088734155,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 20426
    },
    {
      "epoch": 4.862941141462834,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 20427
    },
    {
      "epoch": 4.863179194191513,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 20428
    },
    {
      "epoch": 4.863417246920193,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 20429
    },
    {
      "epoch": 4.863655299648872,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 20430
    },
    {
      "epoch": 4.863893352377552,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 20431
    },
    {
      "epoch": 4.864131405106231,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 20432
    },
    {
      "epoch": 4.86436945783491,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 20433
    },
    {
      "epoch": 4.86460751056359,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 20434
    },
    {
      "epoch": 4.86484556329227,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 20435
    },
    {
      "epoch": 4.865083616020948,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 20436
    },
    {
      "epoch": 4.865321668749628,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 20437
    },
    {
      "epoch": 4.8655597214783075,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 20438
    },
    {
      "epoch": 4.865797774206987,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 20439
    },
    {
      "epoch": 4.866035826935667,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 20440
    },
    {
      "epoch": 4.866273879664345,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3767,
      "step": 20441
    },
    {
      "epoch": 4.866511932393025,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 20442
    },
    {
      "epoch": 4.866749985121705,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 20443
    },
    {
      "epoch": 4.866988037850383,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 20444
    },
    {
      "epoch": 4.867226090579063,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 20445
    },
    {
      "epoch": 4.867464143307743,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 20446
    },
    {
      "epoch": 4.867702196036422,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 20447
    },
    {
      "epoch": 4.867940248765102,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 20448
    },
    {
      "epoch": 4.8681783014937805,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 20449
    },
    {
      "epoch": 4.86841635422246,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 20450
    },
    {
      "epoch": 4.86865440695114,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 20451
    },
    {
      "epoch": 4.868892459679819,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 20452
    },
    {
      "epoch": 4.869130512408498,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 20453
    },
    {
      "epoch": 4.869368565137178,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 20454
    },
    {
      "epoch": 4.869606617865857,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 20455
    },
    {
      "epoch": 4.869844670594537,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 20456
    },
    {
      "epoch": 4.8700827233232165,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 20457
    },
    {
      "epoch": 4.870320776051895,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 20458
    },
    {
      "epoch": 4.870558828780575,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 20459
    },
    {
      "epoch": 4.870796881509254,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 20460
    },
    {
      "epoch": 4.871034934237934,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3295,
      "step": 20461
    },
    {
      "epoch": 4.871272986966613,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 20462
    },
    {
      "epoch": 4.871511039695292,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 20463
    },
    {
      "epoch": 4.871749092423972,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.5087,
      "step": 20464
    },
    {
      "epoch": 4.8719871451526515,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 20465
    },
    {
      "epoch": 4.872225197881331,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 20466
    },
    {
      "epoch": 4.87246325061001,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 20467
    },
    {
      "epoch": 4.8727013033386894,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 20468
    },
    {
      "epoch": 4.872939356067369,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 20469
    },
    {
      "epoch": 4.873177408796049,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 20470
    },
    {
      "epoch": 4.873415461524727,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 20471
    },
    {
      "epoch": 4.873653514253407,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.227,
      "step": 20472
    },
    {
      "epoch": 4.873891566982087,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 20473
    },
    {
      "epoch": 4.874129619710766,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 20474
    },
    {
      "epoch": 4.874367672439446,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 20475
    },
    {
      "epoch": 4.8746057251681245,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 20476
    },
    {
      "epoch": 4.874843777896804,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 20477
    },
    {
      "epoch": 4.875081830625484,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 20478
    },
    {
      "epoch": 4.875319883354163,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 20479
    },
    {
      "epoch": 4.875557936082842,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 20480
    },
    {
      "epoch": 4.875795988811522,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 20481
    },
    {
      "epoch": 4.876034041540201,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4158,
      "step": 20482
    },
    {
      "epoch": 4.876272094268881,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4842,
      "step": 20483
    },
    {
      "epoch": 4.87651014699756,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 20484
    },
    {
      "epoch": 4.876748199726239,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 20485
    },
    {
      "epoch": 4.876986252454919,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 20486
    },
    {
      "epoch": 4.877224305183598,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 20487
    },
    {
      "epoch": 4.877462357912277,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 20488
    },
    {
      "epoch": 4.877700410640957,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 20489
    },
    {
      "epoch": 4.877938463369636,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 20490
    },
    {
      "epoch": 4.878176516098316,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 20491
    },
    {
      "epoch": 4.8784145688269955,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 20492
    },
    {
      "epoch": 4.878652621555674,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 20493
    },
    {
      "epoch": 4.878890674284354,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 20494
    },
    {
      "epoch": 4.8791287270130335,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 20495
    },
    {
      "epoch": 4.879366779741713,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 20496
    },
    {
      "epoch": 4.879604832470392,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 20497
    },
    {
      "epoch": 4.879842885199071,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 20498
    },
    {
      "epoch": 4.880080937927751,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3338,
      "step": 20499
    },
    {
      "epoch": 4.880318990656431,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 20500
    },
    {
      "epoch": 4.880318990656431,
      "eval_runtime": 0.5196,
      "eval_samples_per_second": 1.925,
      "eval_steps_per_second": 1.925,
      "step": 20500
    },
    {
      "epoch": 4.88055704338511,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3144,
      "step": 20501
    },
    {
      "epoch": 4.880795096113789,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 20502
    },
    {
      "epoch": 4.8810331488424685,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 20503
    },
    {
      "epoch": 4.881271201571148,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 20504
    },
    {
      "epoch": 4.881509254299828,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 20505
    },
    {
      "epoch": 4.8817473070285065,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 20506
    },
    {
      "epoch": 4.881985359757186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 20507
    },
    {
      "epoch": 4.882223412485866,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 20508
    },
    {
      "epoch": 4.882461465214545,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 20509
    },
    {
      "epoch": 4.882699517943225,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 20510
    },
    {
      "epoch": 4.882937570671904,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 20511
    },
    {
      "epoch": 4.883175623400583,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 20512
    },
    {
      "epoch": 4.883413676129263,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3884,
      "step": 20513
    },
    {
      "epoch": 4.883651728857942,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 20514
    },
    {
      "epoch": 4.883889781586621,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 20515
    },
    {
      "epoch": 4.884127834315301,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 20516
    },
    {
      "epoch": 4.88436588704398,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 20517
    },
    {
      "epoch": 4.88460393977266,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5972,
      "step": 20518
    },
    {
      "epoch": 4.884841992501339,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 20519
    },
    {
      "epoch": 4.885080045230018,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 20520
    },
    {
      "epoch": 4.885318097958698,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 20521
    },
    {
      "epoch": 4.8855561506873775,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 20522
    },
    {
      "epoch": 4.885794203416056,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.7028,
      "step": 20523
    },
    {
      "epoch": 4.886032256144736,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5014,
      "step": 20524
    },
    {
      "epoch": 4.886270308873415,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 20525
    },
    {
      "epoch": 4.886508361602095,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 20526
    },
    {
      "epoch": 4.886746414330775,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 20527
    },
    {
      "epoch": 4.886984467059453,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 20528
    },
    {
      "epoch": 4.887222519788133,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 20529
    },
    {
      "epoch": 4.8874605725168125,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 20530
    },
    {
      "epoch": 4.887698625245492,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4535,
      "step": 20531
    },
    {
      "epoch": 4.887936677974171,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 20532
    },
    {
      "epoch": 4.8881747307028505,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 20533
    },
    {
      "epoch": 4.88841278343153,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4097,
      "step": 20534
    },
    {
      "epoch": 4.88865083616021,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 20535
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 20536
    },
    {
      "epoch": 4.889126941617568,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 20537
    },
    {
      "epoch": 4.889364994346248,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 20538
    },
    {
      "epoch": 4.889603047074927,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 20539
    },
    {
      "epoch": 4.889841099803607,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 20540
    },
    {
      "epoch": 4.8900791525322855,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 20541
    },
    {
      "epoch": 4.890317205260965,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 20542
    },
    {
      "epoch": 4.890555257989645,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2037,
      "step": 20543
    },
    {
      "epoch": 4.890793310718324,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 20544
    },
    {
      "epoch": 4.891031363447004,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 20545
    },
    {
      "epoch": 4.891269416175683,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 20546
    },
    {
      "epoch": 4.891507468904362,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 20547
    },
    {
      "epoch": 4.891745521633042,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 20548
    },
    {
      "epoch": 4.8919835743617215,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 20549
    },
    {
      "epoch": 4.8922216270904,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 20550
    },
    {
      "epoch": 4.89245967981908,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 20551
    },
    {
      "epoch": 4.892697732547759,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 20552
    },
    {
      "epoch": 4.892935785276439,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 20553
    },
    {
      "epoch": 4.893173838005119,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 20554
    },
    {
      "epoch": 4.893411890733797,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 20555
    },
    {
      "epoch": 4.893649943462477,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 20556
    },
    {
      "epoch": 4.8938879961911566,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 20557
    },
    {
      "epoch": 4.894126048919835,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 20558
    },
    {
      "epoch": 4.894364101648515,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 20559
    },
    {
      "epoch": 4.8946021543771945,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.5054,
      "step": 20560
    },
    {
      "epoch": 4.894840207105874,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 20561
    },
    {
      "epoch": 4.895078259834554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 20562
    },
    {
      "epoch": 4.895316312563232,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 20563
    },
    {
      "epoch": 4.895554365291912,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 20564
    },
    {
      "epoch": 4.895792418020592,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 20565
    },
    {
      "epoch": 4.896030470749271,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 20566
    },
    {
      "epoch": 4.89626852347795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 20567
    },
    {
      "epoch": 4.8965065762066295,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3427,
      "step": 20568
    },
    {
      "epoch": 4.896744628935309,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 20569
    },
    {
      "epoch": 4.896982681663989,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 20570
    },
    {
      "epoch": 4.897220734392668,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5951,
      "step": 20571
    },
    {
      "epoch": 4.897458787121347,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 20572
    },
    {
      "epoch": 4.897696839850027,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 20573
    },
    {
      "epoch": 4.897934892578706,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 20574
    },
    {
      "epoch": 4.898172945307386,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 20575
    },
    {
      "epoch": 4.898410998036065,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 20576
    },
    {
      "epoch": 4.898649050764744,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 20577
    },
    {
      "epoch": 4.898887103493424,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3897,
      "step": 20578
    },
    {
      "epoch": 4.899125156222103,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 20579
    },
    {
      "epoch": 4.899363208950783,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 20580
    },
    {
      "epoch": 4.899601261679462,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 20581
    },
    {
      "epoch": 4.899839314408141,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3709,
      "step": 20582
    },
    {
      "epoch": 4.900077367136821,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 20583
    },
    {
      "epoch": 4.900315419865501,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 20584
    },
    {
      "epoch": 4.900553472594179,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 20585
    },
    {
      "epoch": 4.900791525322859,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 20586
    },
    {
      "epoch": 4.9010295780515385,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 20587
    },
    {
      "epoch": 4.901267630780218,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 20588
    },
    {
      "epoch": 4.901505683508898,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 20589
    },
    {
      "epoch": 4.901743736237576,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2496,
      "step": 20590
    },
    {
      "epoch": 4.901981788966256,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 20591
    },
    {
      "epoch": 4.902219841694936,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3937,
      "step": 20592
    },
    {
      "epoch": 4.902457894423614,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 20593
    },
    {
      "epoch": 4.902695947152294,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3455,
      "step": 20594
    },
    {
      "epoch": 4.902933999880974,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 20595
    },
    {
      "epoch": 4.903172052609653,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 20596
    },
    {
      "epoch": 4.903410105338333,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 20597
    },
    {
      "epoch": 4.9036481580670115,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 20598
    },
    {
      "epoch": 4.903886210795691,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 20599
    },
    {
      "epoch": 4.904124263524371,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 20600
    },
    {
      "epoch": 4.904124263524371,
      "eval_runtime": 0.5166,
      "eval_samples_per_second": 1.936,
      "eval_steps_per_second": 1.936,
      "step": 20600
    },
    {
      "epoch": 4.90436231625305,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 20601
    },
    {
      "epoch": 4.904600368981729,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 20602
    },
    {
      "epoch": 4.904838421710409,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 20603
    },
    {
      "epoch": 4.905076474439088,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 20604
    },
    {
      "epoch": 4.905314527167768,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 20605
    },
    {
      "epoch": 4.905552579896447,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 20606
    },
    {
      "epoch": 4.905790632625126,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 20607
    },
    {
      "epoch": 4.906028685353806,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 20608
    },
    {
      "epoch": 4.906266738082485,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 20609
    },
    {
      "epoch": 4.906504790811165,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 20610
    },
    {
      "epoch": 4.906742843539844,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 20611
    },
    {
      "epoch": 4.906980896268523,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 20612
    },
    {
      "epoch": 4.907218948997203,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 20613
    },
    {
      "epoch": 4.9074570017258825,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.335,
      "step": 20614
    },
    {
      "epoch": 4.907695054454562,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 20615
    },
    {
      "epoch": 4.907933107183241,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 20616
    },
    {
      "epoch": 4.90817115991192,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9081,
      "step": 20617
    },
    {
      "epoch": 4.9084092126406,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 20618
    },
    {
      "epoch": 4.90864726536928,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.5724,
      "step": 20619
    },
    {
      "epoch": 4.908885318097958,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 20620
    },
    {
      "epoch": 4.909123370826638,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3353,
      "step": 20621
    },
    {
      "epoch": 4.909361423555318,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 20622
    },
    {
      "epoch": 4.909599476283997,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4509,
      "step": 20623
    },
    {
      "epoch": 4.909837529012677,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 20624
    },
    {
      "epoch": 4.9100755817413555,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2798,
      "step": 20625
    },
    {
      "epoch": 4.910313634470035,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 20626
    },
    {
      "epoch": 4.910551687198715,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 20627
    },
    {
      "epoch": 4.910789739927393,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 20628
    },
    {
      "epoch": 4.911027792656073,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 20629
    },
    {
      "epoch": 4.911265845384753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3722,
      "step": 20630
    },
    {
      "epoch": 4.911503898113432,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 20631
    },
    {
      "epoch": 4.911741950842112,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4723,
      "step": 20632
    },
    {
      "epoch": 4.911980003570791,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5968,
      "step": 20633
    },
    {
      "epoch": 4.91221805629947,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7824,
      "step": 20634
    },
    {
      "epoch": 4.91245610902815,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 20635
    },
    {
      "epoch": 4.912694161756829,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 20636
    },
    {
      "epoch": 4.912932214485508,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 20637
    },
    {
      "epoch": 4.913170267214188,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 20638
    },
    {
      "epoch": 4.913408319942867,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 20639
    },
    {
      "epoch": 4.913646372671547,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 20640
    },
    {
      "epoch": 4.9138844254002265,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 20641
    },
    {
      "epoch": 4.914122478128905,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 20642
    },
    {
      "epoch": 4.914360530857585,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 20643
    },
    {
      "epoch": 4.9145985835862644,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 20644
    },
    {
      "epoch": 4.914836636314944,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.903,
      "step": 20645
    },
    {
      "epoch": 4.915074689043623,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 20646
    },
    {
      "epoch": 4.915312741772302,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 20647
    },
    {
      "epoch": 4.915550794500982,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 20648
    },
    {
      "epoch": 4.915788847229662,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 20649
    },
    {
      "epoch": 4.916026899958341,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3872,
      "step": 20650
    },
    {
      "epoch": 4.91626495268702,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 20651
    },
    {
      "epoch": 4.9165030054156995,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 20652
    },
    {
      "epoch": 4.916741058144379,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 20653
    },
    {
      "epoch": 4.916979110873059,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 20654
    },
    {
      "epoch": 4.917217163601737,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 20655
    },
    {
      "epoch": 4.917455216330417,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5018,
      "step": 20656
    },
    {
      "epoch": 4.917693269059097,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 20657
    },
    {
      "epoch": 4.917931321787776,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 20658
    },
    {
      "epoch": 4.918169374516456,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 20659
    },
    {
      "epoch": 4.918407427245135,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 20660
    },
    {
      "epoch": 4.918645479973814,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 20661
    },
    {
      "epoch": 4.918883532702494,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2649,
      "step": 20662
    },
    {
      "epoch": 4.9191215854311725,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 20663
    },
    {
      "epoch": 4.919359638159852,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 20664
    },
    {
      "epoch": 4.919597690888532,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 20665
    },
    {
      "epoch": 4.919835743617211,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 20666
    },
    {
      "epoch": 4.920073796345891,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 20667
    },
    {
      "epoch": 4.92031184907457,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 20668
    },
    {
      "epoch": 4.920549901803249,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 20669
    },
    {
      "epoch": 4.920787954531929,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 20670
    },
    {
      "epoch": 4.9210260072606085,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 20671
    },
    {
      "epoch": 4.921264059989287,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 20672
    },
    {
      "epoch": 4.921502112717967,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 20673
    },
    {
      "epoch": 4.921740165446646,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 20674
    },
    {
      "epoch": 4.921978218175326,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 20675
    },
    {
      "epoch": 4.922216270904006,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 20676
    },
    {
      "epoch": 4.922454323632684,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 20677
    },
    {
      "epoch": 4.922692376361364,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 20678
    },
    {
      "epoch": 4.9229304290900435,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3154,
      "step": 20679
    },
    {
      "epoch": 4.923168481818723,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3136,
      "step": 20680
    },
    {
      "epoch": 4.923406534547402,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 20681
    },
    {
      "epoch": 4.9236445872760815,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 20682
    },
    {
      "epoch": 4.923882640004761,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 20683
    },
    {
      "epoch": 4.924120692733441,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3351,
      "step": 20684
    },
    {
      "epoch": 4.92435874546212,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 20685
    },
    {
      "epoch": 4.924596798190799,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 20686
    },
    {
      "epoch": 4.924834850919479,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 20687
    },
    {
      "epoch": 4.925072903648158,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 20688
    },
    {
      "epoch": 4.925310956376838,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 20689
    },
    {
      "epoch": 4.9255490091055165,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 20690
    },
    {
      "epoch": 4.925787061834196,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7903,
      "step": 20691
    },
    {
      "epoch": 4.926025114562876,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 20692
    },
    {
      "epoch": 4.926263167291555,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 20693
    },
    {
      "epoch": 4.926501220020235,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 20694
    },
    {
      "epoch": 4.926739272748914,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3865,
      "step": 20695
    },
    {
      "epoch": 4.926977325477593,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4373,
      "step": 20696
    },
    {
      "epoch": 4.927215378206273,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 20697
    },
    {
      "epoch": 4.9274534309349525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 20698
    },
    {
      "epoch": 4.927691483663631,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 20699
    },
    {
      "epoch": 4.927929536392311,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 20700
    },
    {
      "epoch": 4.927929536392311,
      "eval_runtime": 0.5276,
      "eval_samples_per_second": 1.895,
      "eval_steps_per_second": 1.895,
      "step": 20700
    },
    {
      "epoch": 4.92816758912099,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 20701
    },
    {
      "epoch": 4.92840564184967,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 20702
    },
    {
      "epoch": 4.928643694578349,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 20703
    },
    {
      "epoch": 4.928881747307028,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 20704
    },
    {
      "epoch": 4.929119800035708,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 20705
    },
    {
      "epoch": 4.9293578527643875,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 20706
    },
    {
      "epoch": 4.929595905493066,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 20707
    },
    {
      "epoch": 4.929833958221746,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 20708
    },
    {
      "epoch": 4.9300720109504255,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.672,
      "step": 20709
    },
    {
      "epoch": 4.930310063679105,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.6381,
      "step": 20710
    },
    {
      "epoch": 4.930548116407785,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 20711
    },
    {
      "epoch": 4.930786169136463,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 20712
    },
    {
      "epoch": 4.931024221865143,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 20713
    },
    {
      "epoch": 4.931262274593823,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2391,
      "step": 20714
    },
    {
      "epoch": 4.931500327322502,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.351,
      "step": 20715
    },
    {
      "epoch": 4.931738380051181,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 20716
    },
    {
      "epoch": 4.9319764327798605,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 20717
    },
    {
      "epoch": 4.93221448550854,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 20718
    },
    {
      "epoch": 4.93245253823722,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 20719
    },
    {
      "epoch": 4.932690590965899,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2477,
      "step": 20720
    },
    {
      "epoch": 4.932928643694578,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 20721
    },
    {
      "epoch": 4.933166696423258,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 20722
    },
    {
      "epoch": 4.933404749151937,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 20723
    },
    {
      "epoch": 4.933642801880617,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 20724
    },
    {
      "epoch": 4.933880854609296,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 20725
    },
    {
      "epoch": 4.934118907337975,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 20726
    },
    {
      "epoch": 4.934356960066655,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.5261,
      "step": 20727
    },
    {
      "epoch": 4.934595012795334,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 20728
    },
    {
      "epoch": 4.934833065524014,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 20729
    },
    {
      "epoch": 4.935071118252693,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 20730
    },
    {
      "epoch": 4.935309170981372,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 20731
    },
    {
      "epoch": 4.935547223710052,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 20732
    },
    {
      "epoch": 4.9357852764387316,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 20733
    },
    {
      "epoch": 4.93602332916741,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 20734
    },
    {
      "epoch": 4.93626138189609,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2919,
      "step": 20735
    },
    {
      "epoch": 4.9364994346247695,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 20736
    },
    {
      "epoch": 4.936737487353449,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 20737
    },
    {
      "epoch": 4.936975540082129,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 20738
    },
    {
      "epoch": 4.937213592810807,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 20739
    },
    {
      "epoch": 4.937451645539487,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 20740
    },
    {
      "epoch": 4.937689698268167,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 20741
    },
    {
      "epoch": 4.937927750996845,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 20742
    },
    {
      "epoch": 4.938165803725525,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 20743
    },
    {
      "epoch": 4.9384038564542045,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 20744
    },
    {
      "epoch": 4.938641909182884,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.4392,
      "step": 20745
    },
    {
      "epoch": 4.938879961911564,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 20746
    },
    {
      "epoch": 4.9391180146402425,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 20747
    },
    {
      "epoch": 4.939356067368922,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2804,
      "step": 20748
    },
    {
      "epoch": 4.939594120097602,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 20749
    },
    {
      "epoch": 4.939832172826281,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2872,
      "step": 20750
    },
    {
      "epoch": 4.94007022555496,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 20751
    },
    {
      "epoch": 4.94030827828364,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 20752
    },
    {
      "epoch": 4.940546331012319,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7836,
      "step": 20753
    },
    {
      "epoch": 4.940784383740999,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 20754
    },
    {
      "epoch": 4.941022436469678,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 20755
    },
    {
      "epoch": 4.941260489198357,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 20756
    },
    {
      "epoch": 4.941498541927037,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 20757
    },
    {
      "epoch": 4.941736594655716,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 20758
    },
    {
      "epoch": 4.941974647384396,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 20759
    },
    {
      "epoch": 4.942212700113075,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 20760
    },
    {
      "epoch": 4.942450752841754,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 20761
    },
    {
      "epoch": 4.942688805570434,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 20762
    },
    {
      "epoch": 4.9429268582991135,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 20763
    },
    {
      "epoch": 4.943164911027793,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 20764
    },
    {
      "epoch": 4.943402963756472,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 20765
    },
    {
      "epoch": 4.943641016485151,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 20766
    },
    {
      "epoch": 4.943879069213831,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 20767
    },
    {
      "epoch": 4.944117121942511,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 20768
    },
    {
      "epoch": 4.944355174671189,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 20769
    },
    {
      "epoch": 4.944593227399869,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 20770
    },
    {
      "epoch": 4.944831280128549,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 20771
    },
    {
      "epoch": 4.945069332857228,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 20772
    },
    {
      "epoch": 4.945307385585908,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3686,
      "step": 20773
    },
    {
      "epoch": 4.9455454383145865,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 20774
    },
    {
      "epoch": 4.945783491043266,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3258,
      "step": 20775
    },
    {
      "epoch": 4.946021543771946,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 20776
    },
    {
      "epoch": 4.946259596500624,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 20777
    },
    {
      "epoch": 4.946497649229304,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 20778
    },
    {
      "epoch": 4.946735701957984,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 20779
    },
    {
      "epoch": 4.946973754686663,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8588,
      "step": 20780
    },
    {
      "epoch": 4.947211807415343,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 20781
    },
    {
      "epoch": 4.9474498601440215,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 20782
    },
    {
      "epoch": 4.947687912872701,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 20783
    },
    {
      "epoch": 4.947925965601381,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 20784
    },
    {
      "epoch": 4.94816401833006,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 20785
    },
    {
      "epoch": 4.948402071058739,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 20786
    },
    {
      "epoch": 4.948640123787419,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 20787
    },
    {
      "epoch": 4.948878176516098,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 20788
    },
    {
      "epoch": 4.949116229244778,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 20789
    },
    {
      "epoch": 4.9493542819734575,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.6957,
      "step": 20790
    },
    {
      "epoch": 4.949592334702136,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 20791
    },
    {
      "epoch": 4.949830387430816,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2896,
      "step": 20792
    },
    {
      "epoch": 4.950068440159495,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7091,
      "step": 20793
    },
    {
      "epoch": 4.950306492888175,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 20794
    },
    {
      "epoch": 4.950544545616854,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 20795
    },
    {
      "epoch": 4.950782598345533,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 20796
    },
    {
      "epoch": 4.951020651074213,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 20797
    },
    {
      "epoch": 4.951258703802893,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 20798
    },
    {
      "epoch": 4.951496756531572,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2629,
      "step": 20799
    },
    {
      "epoch": 4.951734809260251,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 20800
    },
    {
      "epoch": 4.951734809260251,
      "eval_runtime": 0.5344,
      "eval_samples_per_second": 1.871,
      "eval_steps_per_second": 1.871,
      "step": 20800
    },
    {
      "epoch": 4.9519728619889305,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 20801
    },
    {
      "epoch": 4.95221091471761,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 20802
    },
    {
      "epoch": 4.95244896744629,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 20803
    },
    {
      "epoch": 4.952687020174968,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 20804
    },
    {
      "epoch": 4.952925072903648,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4342,
      "step": 20805
    },
    {
      "epoch": 4.953163125632328,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 20806
    },
    {
      "epoch": 4.953401178361007,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.295,
      "step": 20807
    },
    {
      "epoch": 4.953639231089687,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 20808
    },
    {
      "epoch": 4.953877283818366,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3367,
      "step": 20809
    },
    {
      "epoch": 4.954115336547045,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 20810
    },
    {
      "epoch": 4.954353389275725,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 20811
    },
    {
      "epoch": 4.9545914420044035,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2487,
      "step": 20812
    },
    {
      "epoch": 4.954829494733083,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7142,
      "step": 20813
    },
    {
      "epoch": 4.955067547461763,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 20814
    },
    {
      "epoch": 4.955305600190442,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 20815
    },
    {
      "epoch": 4.955543652919122,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 20816
    },
    {
      "epoch": 4.955781705647801,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3605,
      "step": 20817
    },
    {
      "epoch": 4.95601975837648,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 20818
    },
    {
      "epoch": 4.95625781110516,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 20819
    },
    {
      "epoch": 4.956495863833839,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 20820
    },
    {
      "epoch": 4.956733916562518,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3245,
      "step": 20821
    },
    {
      "epoch": 4.956971969291198,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 20822
    },
    {
      "epoch": 4.957210022019877,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 20823
    },
    {
      "epoch": 4.957448074748557,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7701,
      "step": 20824
    },
    {
      "epoch": 4.957686127477237,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 20825
    },
    {
      "epoch": 4.957924180205915,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 20826
    },
    {
      "epoch": 4.958162232934595,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2247,
      "step": 20827
    },
    {
      "epoch": 4.9584002856632745,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 20828
    },
    {
      "epoch": 4.958638338391954,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 20829
    },
    {
      "epoch": 4.958876391120633,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 20830
    },
    {
      "epoch": 4.959114443849312,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 20831
    },
    {
      "epoch": 4.959352496577992,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3065,
      "step": 20832
    },
    {
      "epoch": 4.959590549306672,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 20833
    },
    {
      "epoch": 4.959828602035351,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 20834
    },
    {
      "epoch": 4.96006665476403,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8777,
      "step": 20835
    },
    {
      "epoch": 4.96030470749271,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 20836
    },
    {
      "epoch": 4.960542760221389,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 20837
    },
    {
      "epoch": 4.960780812950069,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3482,
      "step": 20838
    },
    {
      "epoch": 4.9610188656787475,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 20839
    },
    {
      "epoch": 4.961256918407427,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9074,
      "step": 20840
    },
    {
      "epoch": 4.961494971136107,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 20841
    },
    {
      "epoch": 4.961733023864786,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8267,
      "step": 20842
    },
    {
      "epoch": 4.961971076593466,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 20843
    },
    {
      "epoch": 4.962209129322145,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 20844
    },
    {
      "epoch": 4.962447182050824,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4209,
      "step": 20845
    },
    {
      "epoch": 4.962685234779504,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 20846
    },
    {
      "epoch": 4.962923287508183,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3091,
      "step": 20847
    },
    {
      "epoch": 4.963161340236862,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 20848
    },
    {
      "epoch": 4.963399392965542,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 20849
    },
    {
      "epoch": 4.963637445694221,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 20850
    },
    {
      "epoch": 4.963875498422901,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8696,
      "step": 20851
    },
    {
      "epoch": 4.96411355115158,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 20852
    },
    {
      "epoch": 4.964351603880259,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 20853
    },
    {
      "epoch": 4.964589656608939,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 20854
    },
    {
      "epoch": 4.9648277093376185,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 20855
    },
    {
      "epoch": 4.965065762066297,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 20856
    },
    {
      "epoch": 4.965303814794977,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 20857
    },
    {
      "epoch": 4.9655418675236564,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 20858
    },
    {
      "epoch": 4.965779920252336,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 20859
    },
    {
      "epoch": 4.966017972981016,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 20860
    },
    {
      "epoch": 4.966256025709694,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 20861
    },
    {
      "epoch": 4.966494078438374,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 20862
    },
    {
      "epoch": 4.966732131167054,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2619,
      "step": 20863
    },
    {
      "epoch": 4.966970183895733,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 20864
    },
    {
      "epoch": 4.967208236624412,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 20865
    },
    {
      "epoch": 4.9674462893530915,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 20866
    },
    {
      "epoch": 4.967684342081771,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 20867
    },
    {
      "epoch": 4.967922394810451,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 20868
    },
    {
      "epoch": 4.96816044753913,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 20869
    },
    {
      "epoch": 4.968398500267809,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.481,
      "step": 20870
    },
    {
      "epoch": 4.968636552996489,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 20871
    },
    {
      "epoch": 4.968874605725168,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 20872
    },
    {
      "epoch": 4.969112658453848,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 20873
    },
    {
      "epoch": 4.969350711182527,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 20874
    },
    {
      "epoch": 4.969588763911206,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 20875
    },
    {
      "epoch": 4.969826816639886,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.4083,
      "step": 20876
    },
    {
      "epoch": 4.970064869368565,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 20877
    },
    {
      "epoch": 4.970302922097245,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 20878
    },
    {
      "epoch": 4.970540974825924,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 20879
    },
    {
      "epoch": 4.970779027554603,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3696,
      "step": 20880
    },
    {
      "epoch": 4.971017080283283,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 20881
    },
    {
      "epoch": 4.9712551330119625,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 20882
    },
    {
      "epoch": 4.971493185740641,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 20883
    },
    {
      "epoch": 4.971731238469321,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 20884
    },
    {
      "epoch": 4.9719692911980005,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 20885
    },
    {
      "epoch": 4.97220734392668,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3012,
      "step": 20886
    },
    {
      "epoch": 4.972445396655359,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 20887
    },
    {
      "epoch": 4.972683449384038,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 20888
    },
    {
      "epoch": 4.972921502112718,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 20889
    },
    {
      "epoch": 4.973159554841398,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 20890
    },
    {
      "epoch": 4.973397607570076,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 20891
    },
    {
      "epoch": 4.973635660298756,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 20892
    },
    {
      "epoch": 4.9738737130274355,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 20893
    },
    {
      "epoch": 4.974111765756115,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 20894
    },
    {
      "epoch": 4.974349818484795,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 20895
    },
    {
      "epoch": 4.9745878712134735,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 20896
    },
    {
      "epoch": 4.974825923942153,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 20897
    },
    {
      "epoch": 4.975063976670833,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 20898
    },
    {
      "epoch": 4.975302029399512,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 20899
    },
    {
      "epoch": 4.975540082128191,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4743,
      "step": 20900
    },
    {
      "epoch": 4.975540082128191,
      "eval_runtime": 0.5314,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 20900
    },
    {
      "epoch": 4.975778134856871,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 20901
    },
    {
      "epoch": 4.97601618758555,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 20902
    },
    {
      "epoch": 4.97625424031423,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 20903
    },
    {
      "epoch": 4.976492293042909,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 20904
    },
    {
      "epoch": 4.976730345771588,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 20905
    },
    {
      "epoch": 4.976968398500268,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 20906
    },
    {
      "epoch": 4.977206451228947,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 20907
    },
    {
      "epoch": 4.977444503957627,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 20908
    },
    {
      "epoch": 4.977682556686306,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 20909
    },
    {
      "epoch": 4.977920609414985,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9388,
      "step": 20910
    },
    {
      "epoch": 4.978158662143665,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 20911
    },
    {
      "epoch": 4.9783967148723445,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 20912
    },
    {
      "epoch": 4.978634767601024,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 20913
    },
    {
      "epoch": 4.978872820329703,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 20914
    },
    {
      "epoch": 4.979110873058382,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 20915
    },
    {
      "epoch": 4.979348925787062,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 20916
    },
    {
      "epoch": 4.979586978515742,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 20917
    },
    {
      "epoch": 4.97982503124442,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 20918
    },
    {
      "epoch": 4.9800630839731,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 20919
    },
    {
      "epoch": 4.9803011367017795,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 20920
    },
    {
      "epoch": 4.980539189430459,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 20921
    },
    {
      "epoch": 4.980777242159138,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 20922
    },
    {
      "epoch": 4.9810152948878175,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 20923
    },
    {
      "epoch": 4.981253347616497,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 20924
    },
    {
      "epoch": 4.981491400345177,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 20925
    },
    {
      "epoch": 4.981729453073855,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 20926
    },
    {
      "epoch": 4.981967505802535,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 20927
    },
    {
      "epoch": 4.982205558531215,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 20928
    },
    {
      "epoch": 4.982443611259894,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 20929
    },
    {
      "epoch": 4.982681663988574,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 20930
    },
    {
      "epoch": 4.9829197167172525,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 20931
    },
    {
      "epoch": 4.983157769445932,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2492,
      "step": 20932
    },
    {
      "epoch": 4.983395822174612,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 20933
    },
    {
      "epoch": 4.983633874903291,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 20934
    },
    {
      "epoch": 4.98387192763197,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 20935
    },
    {
      "epoch": 4.98410998036065,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 20936
    },
    {
      "epoch": 4.984348033089329,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 20937
    },
    {
      "epoch": 4.984586085818009,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 20938
    },
    {
      "epoch": 4.9848241385466885,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 20939
    },
    {
      "epoch": 4.985062191275367,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 20940
    },
    {
      "epoch": 4.985300244004047,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 20941
    },
    {
      "epoch": 4.985538296732726,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 20942
    },
    {
      "epoch": 4.985776349461406,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3486,
      "step": 20943
    },
    {
      "epoch": 4.986014402190085,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 20944
    },
    {
      "epoch": 4.986252454918764,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 20945
    },
    {
      "epoch": 4.986490507647444,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 20946
    },
    {
      "epoch": 4.9867285603761236,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3187,
      "step": 20947
    },
    {
      "epoch": 4.986966613104803,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 20948
    },
    {
      "epoch": 4.987204665833482,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 20949
    },
    {
      "epoch": 4.9874427185621615,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 20950
    },
    {
      "epoch": 4.987680771290841,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 20951
    },
    {
      "epoch": 4.987918824019521,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 20952
    },
    {
      "epoch": 4.988156876748199,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 20953
    },
    {
      "epoch": 4.988394929476879,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 20954
    },
    {
      "epoch": 4.988632982205559,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 20955
    },
    {
      "epoch": 4.988871034934238,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 20956
    },
    {
      "epoch": 4.989109087662918,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 20957
    },
    {
      "epoch": 4.9893471403915965,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 20958
    },
    {
      "epoch": 4.989585193120276,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2495,
      "step": 20959
    },
    {
      "epoch": 4.989823245848956,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2305,
      "step": 20960
    },
    {
      "epoch": 4.9900612985776345,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 20961
    },
    {
      "epoch": 4.990299351306314,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 20962
    },
    {
      "epoch": 4.990537404034994,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 20963
    },
    {
      "epoch": 4.990775456763673,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 20964
    },
    {
      "epoch": 4.991013509492353,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3943,
      "step": 20965
    },
    {
      "epoch": 4.991251562221032,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 20966
    },
    {
      "epoch": 4.991489614949711,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7552,
      "step": 20967
    },
    {
      "epoch": 4.991727667678391,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 20968
    },
    {
      "epoch": 4.99196572040707,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 20969
    },
    {
      "epoch": 4.992203773135749,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 20970
    },
    {
      "epoch": 4.992441825864429,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 20971
    },
    {
      "epoch": 4.992679878593108,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 20972
    },
    {
      "epoch": 4.992917931321788,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 20973
    },
    {
      "epoch": 4.993155984050468,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 20974
    },
    {
      "epoch": 4.993394036779146,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 20975
    },
    {
      "epoch": 4.993632089507826,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 20976
    },
    {
      "epoch": 4.9938701422365055,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 20977
    },
    {
      "epoch": 4.994108194965185,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3291,
      "step": 20978
    },
    {
      "epoch": 4.994346247693864,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 20979
    },
    {
      "epoch": 4.994584300422543,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 20980
    },
    {
      "epoch": 4.994822353151223,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 20981
    },
    {
      "epoch": 4.995060405879903,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 20982
    },
    {
      "epoch": 4.995298458608582,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 20983
    },
    {
      "epoch": 4.995536511337261,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 20984
    },
    {
      "epoch": 4.995774564065941,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 20985
    },
    {
      "epoch": 4.99601261679462,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 20986
    },
    {
      "epoch": 4.9962506695233,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 20987
    },
    {
      "epoch": 4.9964887222519785,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 20988
    },
    {
      "epoch": 4.996726774980658,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 20989
    },
    {
      "epoch": 4.996964827709338,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 20990
    },
    {
      "epoch": 4.997202880438017,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.4292,
      "step": 20991
    },
    {
      "epoch": 4.997440933166697,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 20992
    },
    {
      "epoch": 4.997678985895376,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 20993
    },
    {
      "epoch": 4.997917038624055,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 20994
    },
    {
      "epoch": 4.998155091352735,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2269,
      "step": 20995
    },
    {
      "epoch": 4.9983931440814136,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 20996
    },
    {
      "epoch": 4.998631196810093,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.406,
      "step": 20997
    },
    {
      "epoch": 4.998869249538773,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 20998
    },
    {
      "epoch": 4.999107302267452,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 20999
    },
    {
      "epoch": 4.999345354996132,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 21000
    },
    {
      "epoch": 4.999345354996132,
      "eval_runtime": 0.529,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 21000
    },
    {
      "epoch": 4.999583407724811,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 21001
    },
    {
      "epoch": 4.99982146045349,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 21002
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 21003
    },
    {
      "epoch": 5.00023805272868,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 21004
    },
    {
      "epoch": 5.000476105457359,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 21005
    },
    {
      "epoch": 5.000714158186038,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 21006
    },
    {
      "epoch": 5.0009522109147175,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 21007
    },
    {
      "epoch": 5.001190263643397,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 21008
    },
    {
      "epoch": 5.001428316372077,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 21009
    },
    {
      "epoch": 5.0016663691007555,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 21010
    },
    {
      "epoch": 5.001904421829435,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 21011
    },
    {
      "epoch": 5.002142474558115,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 21012
    },
    {
      "epoch": 5.002380527286794,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 21013
    },
    {
      "epoch": 5.002618580015473,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 21014
    },
    {
      "epoch": 5.002856632744153,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 21015
    },
    {
      "epoch": 5.003094685472832,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 21016
    },
    {
      "epoch": 5.003332738201512,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 21017
    },
    {
      "epoch": 5.003570790930191,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 21018
    },
    {
      "epoch": 5.00380884365887,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 21019
    },
    {
      "epoch": 5.00404689638755,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 21020
    },
    {
      "epoch": 5.004284949116229,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 21021
    },
    {
      "epoch": 5.004523001844909,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 21022
    },
    {
      "epoch": 5.004761054573588,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 21023
    },
    {
      "epoch": 5.004999107302267,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 21024
    },
    {
      "epoch": 5.005237160030947,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 21025
    },
    {
      "epoch": 5.0054752127596265,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 21026
    },
    {
      "epoch": 5.005713265488306,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 21027
    },
    {
      "epoch": 5.005951318216985,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 21028
    },
    {
      "epoch": 5.006189370945664,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 21029
    },
    {
      "epoch": 5.006427423674344,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 21030
    },
    {
      "epoch": 5.006665476403024,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 21031
    },
    {
      "epoch": 5.006903529131702,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 21032
    },
    {
      "epoch": 5.007141581860382,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 21033
    },
    {
      "epoch": 5.0073796345890615,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 21034
    },
    {
      "epoch": 5.007617687317741,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4762,
      "step": 21035
    },
    {
      "epoch": 5.00785574004642,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 21036
    },
    {
      "epoch": 5.0080937927750995,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.294,
      "step": 21037
    },
    {
      "epoch": 5.008331845503779,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 21038
    },
    {
      "epoch": 5.008569898232459,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 21039
    },
    {
      "epoch": 5.008807950961138,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 21040
    },
    {
      "epoch": 5.009046003689817,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 21041
    },
    {
      "epoch": 5.009284056418497,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 21042
    },
    {
      "epoch": 5.009522109147176,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 21043
    },
    {
      "epoch": 5.009760161875856,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 21044
    },
    {
      "epoch": 5.0099982146045345,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 21045
    },
    {
      "epoch": 5.010236267333214,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 21046
    },
    {
      "epoch": 5.010474320061894,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5945,
      "step": 21047
    },
    {
      "epoch": 5.010712372790573,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 21048
    },
    {
      "epoch": 5.010950425519252,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 21049
    },
    {
      "epoch": 5.011188478247932,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4427,
      "step": 21050
    },
    {
      "epoch": 5.011426530976611,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 21051
    },
    {
      "epoch": 5.011664583705291,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 21052
    },
    {
      "epoch": 5.0119026364339705,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 21053
    },
    {
      "epoch": 5.012140689162649,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 21054
    },
    {
      "epoch": 5.012378741891329,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 21055
    },
    {
      "epoch": 5.012616794620008,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7135,
      "step": 21056
    },
    {
      "epoch": 5.012854847348688,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 21057
    },
    {
      "epoch": 5.013092900077367,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 21058
    },
    {
      "epoch": 5.013330952806046,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4015,
      "step": 21059
    },
    {
      "epoch": 5.013569005534726,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3973,
      "step": 21060
    },
    {
      "epoch": 5.013807058263406,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2547,
      "step": 21061
    },
    {
      "epoch": 5.014045110992085,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 21062
    },
    {
      "epoch": 5.014283163720764,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 21063
    },
    {
      "epoch": 5.0145212164494435,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 21064
    },
    {
      "epoch": 5.014759269178123,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 21065
    },
    {
      "epoch": 5.014997321906803,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 21066
    },
    {
      "epoch": 5.015235374635481,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 21067
    },
    {
      "epoch": 5.015473427364161,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1563,
      "step": 21068
    },
    {
      "epoch": 5.015711480092841,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 21069
    },
    {
      "epoch": 5.01594953282152,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 21070
    },
    {
      "epoch": 5.016187585550199,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 21071
    },
    {
      "epoch": 5.0164256382788786,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 21072
    },
    {
      "epoch": 5.016663691007558,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 21073
    },
    {
      "epoch": 5.016901743736238,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 21074
    },
    {
      "epoch": 5.017139796464917,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 21075
    },
    {
      "epoch": 5.017377849193596,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 21076
    },
    {
      "epoch": 5.017615901922276,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 21077
    },
    {
      "epoch": 5.017853954650955,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 21078
    },
    {
      "epoch": 5.018092007379635,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 21079
    },
    {
      "epoch": 5.018330060108314,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 21080
    },
    {
      "epoch": 5.018568112836993,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 21081
    },
    {
      "epoch": 5.018806165565673,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 21082
    },
    {
      "epoch": 5.019044218294352,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 21083
    },
    {
      "epoch": 5.019282271023032,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 21084
    },
    {
      "epoch": 5.019520323751711,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 21085
    },
    {
      "epoch": 5.01975837648039,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 21086
    },
    {
      "epoch": 5.01999642920907,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 21087
    },
    {
      "epoch": 5.02023448193775,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 21088
    },
    {
      "epoch": 5.020472534666428,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 21089
    },
    {
      "epoch": 5.020710587395108,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 21090
    },
    {
      "epoch": 5.0209486401237875,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 21091
    },
    {
      "epoch": 5.021186692852467,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 21092
    },
    {
      "epoch": 5.021424745581146,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 21093
    },
    {
      "epoch": 5.021662798309825,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 21094
    },
    {
      "epoch": 5.021900851038505,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 21095
    },
    {
      "epoch": 5.022138903767185,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 21096
    },
    {
      "epoch": 5.022376956495864,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.3064,
      "step": 21097
    },
    {
      "epoch": 5.022615009224543,
      "grad_norm": 7.5,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 21098
    },
    {
      "epoch": 5.022853061953223,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 21099
    },
    {
      "epoch": 5.023091114681902,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 21100
    },
    {
      "epoch": 5.023091114681902,
      "eval_runtime": 0.5262,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 21100
    },
    {
      "epoch": 5.023329167410582,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 21101
    },
    {
      "epoch": 5.0235672201392605,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 21102
    },
    {
      "epoch": 5.02380527286794,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 21103
    },
    {
      "epoch": 5.02404332559662,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 21104
    },
    {
      "epoch": 5.024281378325299,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 21105
    },
    {
      "epoch": 5.024519431053978,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 21106
    },
    {
      "epoch": 5.024757483782658,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2355,
      "step": 21107
    },
    {
      "epoch": 5.024995536511337,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8151,
      "step": 21108
    },
    {
      "epoch": 5.025233589240017,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 21109
    },
    {
      "epoch": 5.0254716419686964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 21110
    },
    {
      "epoch": 5.025709694697375,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 21111
    },
    {
      "epoch": 5.025947747426055,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 21112
    },
    {
      "epoch": 5.026185800154734,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3724,
      "step": 21113
    },
    {
      "epoch": 5.026423852883414,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 21114
    },
    {
      "epoch": 5.026661905612093,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 21115
    },
    {
      "epoch": 5.026899958340772,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 21116
    },
    {
      "epoch": 5.027138011069452,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3579,
      "step": 21117
    },
    {
      "epoch": 5.0273760637981315,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 21118
    },
    {
      "epoch": 5.027614116526811,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 21119
    },
    {
      "epoch": 5.02785216925549,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.25,
      "step": 21120
    },
    {
      "epoch": 5.028090221984169,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 21121
    },
    {
      "epoch": 5.028328274712849,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 21122
    },
    {
      "epoch": 5.028566327441529,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 21123
    },
    {
      "epoch": 5.028804380170207,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 21124
    },
    {
      "epoch": 5.029042432898887,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3505,
      "step": 21125
    },
    {
      "epoch": 5.029280485627567,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 21126
    },
    {
      "epoch": 5.029518538356246,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3388,
      "step": 21127
    },
    {
      "epoch": 5.029756591084925,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 21128
    },
    {
      "epoch": 5.0299946438136045,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 21129
    },
    {
      "epoch": 5.030232696542284,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 21130
    },
    {
      "epoch": 5.030470749270964,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 21131
    },
    {
      "epoch": 5.030708801999643,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2781,
      "step": 21132
    },
    {
      "epoch": 5.030946854728322,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 21133
    },
    {
      "epoch": 5.031184907457002,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 21134
    },
    {
      "epoch": 5.031422960185681,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 21135
    },
    {
      "epoch": 5.031661012914361,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 21136
    },
    {
      "epoch": 5.03189906564304,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 21137
    },
    {
      "epoch": 5.032137118371719,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 21138
    },
    {
      "epoch": 5.032375171100399,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2752,
      "step": 21139
    },
    {
      "epoch": 5.032613223829078,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 21140
    },
    {
      "epoch": 5.032851276557757,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 21141
    },
    {
      "epoch": 5.033089329286437,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 21142
    },
    {
      "epoch": 5.033327382015116,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 21143
    },
    {
      "epoch": 5.033565434743796,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 21144
    },
    {
      "epoch": 5.0338034874724755,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 21145
    },
    {
      "epoch": 5.034041540201154,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 21146
    },
    {
      "epoch": 5.034279592929834,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.289,
      "step": 21147
    },
    {
      "epoch": 5.0345176456585135,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 21148
    },
    {
      "epoch": 5.034755698387193,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 21149
    },
    {
      "epoch": 5.034993751115872,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 21150
    },
    {
      "epoch": 5.035231803844551,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 21151
    },
    {
      "epoch": 5.035469856573231,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 21152
    },
    {
      "epoch": 5.035707909301911,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 21153
    },
    {
      "epoch": 5.03594596203059,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 21154
    },
    {
      "epoch": 5.036184014759269,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 21155
    },
    {
      "epoch": 5.0364220674879485,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 21156
    },
    {
      "epoch": 5.036660120216628,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 21157
    },
    {
      "epoch": 5.036898172945308,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 21158
    },
    {
      "epoch": 5.0371362256739864,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 21159
    },
    {
      "epoch": 5.037374278402666,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 21160
    },
    {
      "epoch": 5.037612331131346,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4511,
      "step": 21161
    },
    {
      "epoch": 5.037850383860025,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 21162
    },
    {
      "epoch": 5.038088436588704,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 21163
    },
    {
      "epoch": 5.038326489317384,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 21164
    },
    {
      "epoch": 5.038564542046063,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3951,
      "step": 21165
    },
    {
      "epoch": 5.038802594774743,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 21166
    },
    {
      "epoch": 5.039040647503422,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 21167
    },
    {
      "epoch": 5.039278700232101,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 21168
    },
    {
      "epoch": 5.039516752960781,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 21169
    },
    {
      "epoch": 5.03975480568946,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 21170
    },
    {
      "epoch": 5.03999285841814,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 21171
    },
    {
      "epoch": 5.040230911146819,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 21172
    },
    {
      "epoch": 5.040468963875498,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 21173
    },
    {
      "epoch": 5.040707016604178,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3989,
      "step": 21174
    },
    {
      "epoch": 5.0409450693328575,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 21175
    },
    {
      "epoch": 5.041183122061537,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3199,
      "step": 21176
    },
    {
      "epoch": 5.041421174790216,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 21177
    },
    {
      "epoch": 5.041659227518895,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 21178
    },
    {
      "epoch": 5.041897280247575,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 21179
    },
    {
      "epoch": 5.042135332976255,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 21180
    },
    {
      "epoch": 5.042373385704933,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 21181
    },
    {
      "epoch": 5.042611438433613,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 21182
    },
    {
      "epoch": 5.0428494911622925,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 21183
    },
    {
      "epoch": 5.043087543890972,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 21184
    },
    {
      "epoch": 5.043325596619651,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 21185
    },
    {
      "epoch": 5.0435636493483305,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 21186
    },
    {
      "epoch": 5.04380170207701,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 21187
    },
    {
      "epoch": 5.04403975480569,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 21188
    },
    {
      "epoch": 5.044277807534369,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 21189
    },
    {
      "epoch": 5.044515860263048,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 21190
    },
    {
      "epoch": 5.044753912991728,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2827,
      "step": 21191
    },
    {
      "epoch": 5.044991965720407,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 21192
    },
    {
      "epoch": 5.045230018449087,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 21193
    },
    {
      "epoch": 5.0454680711777655,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 21194
    },
    {
      "epoch": 5.045706123906445,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4096,
      "step": 21195
    },
    {
      "epoch": 5.045944176635125,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 21196
    },
    {
      "epoch": 5.046182229363804,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 21197
    },
    {
      "epoch": 5.046420282092483,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 21198
    },
    {
      "epoch": 5.046658334821163,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 21199
    },
    {
      "epoch": 5.046896387549842,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 21200
    },
    {
      "epoch": 5.046896387549842,
      "eval_runtime": 0.523,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 21200
    },
    {
      "epoch": 5.047134440278522,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 21201
    },
    {
      "epoch": 5.0473724930072015,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4562,
      "step": 21202
    },
    {
      "epoch": 5.04761054573588,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 21203
    },
    {
      "epoch": 5.04784859846456,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 21204
    },
    {
      "epoch": 5.048086651193239,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 21205
    },
    {
      "epoch": 5.048324703921919,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 21206
    },
    {
      "epoch": 5.048562756650598,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 21207
    },
    {
      "epoch": 5.048800809379277,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 21208
    },
    {
      "epoch": 5.049038862107957,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 21209
    },
    {
      "epoch": 5.0492769148366365,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3126,
      "step": 21210
    },
    {
      "epoch": 5.049514967565316,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 21211
    },
    {
      "epoch": 5.049753020293995,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 21212
    },
    {
      "epoch": 5.0499910730226745,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 21213
    },
    {
      "epoch": 5.050229125751354,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 21214
    },
    {
      "epoch": 5.050467178480034,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 21215
    },
    {
      "epoch": 5.050705231208712,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 21216
    },
    {
      "epoch": 5.050943283937392,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 21217
    },
    {
      "epoch": 5.051181336666072,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 21218
    },
    {
      "epoch": 5.051419389394751,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 21219
    },
    {
      "epoch": 5.05165744212343,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 21220
    },
    {
      "epoch": 5.0518954948521095,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 21221
    },
    {
      "epoch": 5.052133547580789,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 21222
    },
    {
      "epoch": 5.052371600309469,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 21223
    },
    {
      "epoch": 5.052609653038148,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 21224
    },
    {
      "epoch": 5.052847705766827,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 21225
    },
    {
      "epoch": 5.053085758495507,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3045,
      "step": 21226
    },
    {
      "epoch": 5.053323811224186,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 21227
    },
    {
      "epoch": 5.053561863952866,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 21228
    },
    {
      "epoch": 5.053799916681545,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 21229
    },
    {
      "epoch": 5.054037969410224,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 21230
    },
    {
      "epoch": 5.054276022138904,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 21231
    },
    {
      "epoch": 5.054514074867583,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 21232
    },
    {
      "epoch": 5.054752127596262,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4042,
      "step": 21233
    },
    {
      "epoch": 5.054990180324942,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 21234
    },
    {
      "epoch": 5.055228233053621,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4089,
      "step": 21235
    },
    {
      "epoch": 5.055466285782301,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 21236
    },
    {
      "epoch": 5.0557043385109806,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 21237
    },
    {
      "epoch": 5.055942391239659,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 21238
    },
    {
      "epoch": 5.056180443968339,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 21239
    },
    {
      "epoch": 5.0564184966970185,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 21240
    },
    {
      "epoch": 5.056656549425698,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 21241
    },
    {
      "epoch": 5.056894602154377,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 21242
    },
    {
      "epoch": 5.057132654883056,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.351,
      "step": 21243
    },
    {
      "epoch": 5.057370707611736,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 21244
    },
    {
      "epoch": 5.057608760340416,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 21245
    },
    {
      "epoch": 5.057846813069095,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 21246
    },
    {
      "epoch": 5.058084865797774,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 21247
    },
    {
      "epoch": 5.0583229185264535,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 21248
    },
    {
      "epoch": 5.058560971255133,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 21249
    },
    {
      "epoch": 5.058799023983813,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 21250
    },
    {
      "epoch": 5.0590370767124915,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 21251
    },
    {
      "epoch": 5.059275129441171,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 21252
    },
    {
      "epoch": 5.059513182169851,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 21253
    },
    {
      "epoch": 5.05975123489853,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 21254
    },
    {
      "epoch": 5.059989287627209,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 21255
    },
    {
      "epoch": 5.060227340355889,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3439,
      "step": 21256
    },
    {
      "epoch": 5.060465393084568,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 21257
    },
    {
      "epoch": 5.060703445813248,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 21258
    },
    {
      "epoch": 5.060941498541927,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 21259
    },
    {
      "epoch": 5.061179551270606,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3091,
      "step": 21260
    },
    {
      "epoch": 5.061417603999286,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 21261
    },
    {
      "epoch": 5.061655656727965,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 21262
    },
    {
      "epoch": 5.061893709456645,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 21263
    },
    {
      "epoch": 5.062131762185324,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 21264
    },
    {
      "epoch": 5.062369814914003,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 21265
    },
    {
      "epoch": 5.062607867642683,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 21266
    },
    {
      "epoch": 5.0628459203713625,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9538,
      "step": 21267
    },
    {
      "epoch": 5.063083973100042,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 21268
    },
    {
      "epoch": 5.063322025828721,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 21269
    },
    {
      "epoch": 5.0635600785574,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 21270
    },
    {
      "epoch": 5.06379813128608,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4455,
      "step": 21271
    },
    {
      "epoch": 5.06403618401476,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 21272
    },
    {
      "epoch": 5.064274236743438,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 21273
    },
    {
      "epoch": 5.064512289472118,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 21274
    },
    {
      "epoch": 5.064750342200798,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 21275
    },
    {
      "epoch": 5.064988394929477,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 21276
    },
    {
      "epoch": 5.065226447658156,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 21277
    },
    {
      "epoch": 5.0654645003868355,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 21278
    },
    {
      "epoch": 5.065702553115515,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 21279
    },
    {
      "epoch": 5.065940605844195,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 21280
    },
    {
      "epoch": 5.066178658572874,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 21281
    },
    {
      "epoch": 5.066416711301553,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 21282
    },
    {
      "epoch": 5.066654764030233,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 21283
    },
    {
      "epoch": 5.066892816758912,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 21284
    },
    {
      "epoch": 5.067130869487592,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 21285
    },
    {
      "epoch": 5.0673689222162706,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2864,
      "step": 21286
    },
    {
      "epoch": 5.06760697494495,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 21287
    },
    {
      "epoch": 5.06784502767363,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 21288
    },
    {
      "epoch": 5.068083080402309,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 21289
    },
    {
      "epoch": 5.068321133130988,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6849,
      "step": 21290
    },
    {
      "epoch": 5.068559185859668,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 21291
    },
    {
      "epoch": 5.068797238588347,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 21292
    },
    {
      "epoch": 5.069035291317027,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 21293
    },
    {
      "epoch": 5.0692733440457065,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 21294
    },
    {
      "epoch": 5.069511396774385,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 21295
    },
    {
      "epoch": 5.069749449503065,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9805,
      "step": 21296
    },
    {
      "epoch": 5.069987502231744,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 21297
    },
    {
      "epoch": 5.070225554960424,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 21298
    },
    {
      "epoch": 5.070463607689103,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 21299
    },
    {
      "epoch": 5.070701660417782,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.407,
      "step": 21300
    },
    {
      "epoch": 5.070701660417782,
      "eval_runtime": 0.533,
      "eval_samples_per_second": 1.876,
      "eval_steps_per_second": 1.876,
      "step": 21300
    },
    {
      "epoch": 5.070939713146462,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 21301
    },
    {
      "epoch": 5.071177765875142,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 21302
    },
    {
      "epoch": 5.071415818603821,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 21303
    },
    {
      "epoch": 5.0716538713325,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 21304
    },
    {
      "epoch": 5.0718919240611795,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 21305
    },
    {
      "epoch": 5.072129976789859,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 21306
    },
    {
      "epoch": 5.072368029518539,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 21307
    },
    {
      "epoch": 5.072606082247217,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 21308
    },
    {
      "epoch": 5.072844134975897,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 21309
    },
    {
      "epoch": 5.073082187704577,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2936,
      "step": 21310
    },
    {
      "epoch": 5.073320240433256,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 21311
    },
    {
      "epoch": 5.073558293161935,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 21312
    },
    {
      "epoch": 5.073796345890615,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 21313
    },
    {
      "epoch": 5.074034398619294,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3789,
      "step": 21314
    },
    {
      "epoch": 5.074272451347974,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 21315
    },
    {
      "epoch": 5.074510504076653,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 21316
    },
    {
      "epoch": 5.074748556805332,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 21317
    },
    {
      "epoch": 5.074986609534012,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 21318
    },
    {
      "epoch": 5.075224662262691,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 21319
    },
    {
      "epoch": 5.075462714991371,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 21320
    },
    {
      "epoch": 5.07570076772005,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 21321
    },
    {
      "epoch": 5.075938820448729,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 21322
    },
    {
      "epoch": 5.076176873177409,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 21323
    },
    {
      "epoch": 5.0764149259060884,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 21324
    },
    {
      "epoch": 5.076652978634767,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 21325
    },
    {
      "epoch": 5.076891031363447,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 21326
    },
    {
      "epoch": 5.077129084092126,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 21327
    },
    {
      "epoch": 5.077367136820806,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 21328
    },
    {
      "epoch": 5.077605189549486,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 21329
    },
    {
      "epoch": 5.077843242278164,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2842,
      "step": 21330
    },
    {
      "epoch": 5.078081295006844,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2341,
      "step": 21331
    },
    {
      "epoch": 5.0783193477355235,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 21332
    },
    {
      "epoch": 5.078557400464203,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 21333
    },
    {
      "epoch": 5.078795453192882,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 21334
    },
    {
      "epoch": 5.079033505921561,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 21335
    },
    {
      "epoch": 5.079271558650241,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 21336
    },
    {
      "epoch": 5.079509611378921,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 21337
    },
    {
      "epoch": 5.0797476641076,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 21338
    },
    {
      "epoch": 5.079985716836279,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 21339
    },
    {
      "epoch": 5.080223769564959,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3598,
      "step": 21340
    },
    {
      "epoch": 5.080461822293638,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1613,
      "step": 21341
    },
    {
      "epoch": 5.080699875022318,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 21342
    },
    {
      "epoch": 5.0809379277509965,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 21343
    },
    {
      "epoch": 5.081175980479676,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 21344
    },
    {
      "epoch": 5.081414033208356,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 21345
    },
    {
      "epoch": 5.081652085937035,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 21346
    },
    {
      "epoch": 5.081890138665714,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 21347
    },
    {
      "epoch": 5.082128191394394,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 21348
    },
    {
      "epoch": 5.082366244123073,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 21349
    },
    {
      "epoch": 5.082604296851753,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 21350
    },
    {
      "epoch": 5.0828423495804325,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 21351
    },
    {
      "epoch": 5.083080402309111,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2569,
      "step": 21352
    },
    {
      "epoch": 5.083318455037791,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 21353
    },
    {
      "epoch": 5.08355650776647,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 21354
    },
    {
      "epoch": 5.08379456049515,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 21355
    },
    {
      "epoch": 5.084032613223829,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 21356
    },
    {
      "epoch": 5.084270665952508,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3276,
      "step": 21357
    },
    {
      "epoch": 5.084508718681188,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 21358
    },
    {
      "epoch": 5.0847467714098675,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 21359
    },
    {
      "epoch": 5.084984824138546,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 21360
    },
    {
      "epoch": 5.085222876867226,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3702,
      "step": 21361
    },
    {
      "epoch": 5.0854609295959055,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2043,
      "step": 21362
    },
    {
      "epoch": 5.085698982324585,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 21363
    },
    {
      "epoch": 5.085937035053265,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 21364
    },
    {
      "epoch": 5.086175087781943,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 21365
    },
    {
      "epoch": 5.086413140510623,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 21366
    },
    {
      "epoch": 5.086651193239303,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 21367
    },
    {
      "epoch": 5.086889245967982,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.32,
      "step": 21368
    },
    {
      "epoch": 5.087127298696661,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 21369
    },
    {
      "epoch": 5.0873653514253405,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 21370
    },
    {
      "epoch": 5.08760340415402,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 21371
    },
    {
      "epoch": 5.0878414568827,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 21372
    },
    {
      "epoch": 5.088079509611379,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2822,
      "step": 21373
    },
    {
      "epoch": 5.088317562340058,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 21374
    },
    {
      "epoch": 5.088555615068738,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 21375
    },
    {
      "epoch": 5.088793667797417,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3864,
      "step": 21376
    },
    {
      "epoch": 5.089031720526097,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 21377
    },
    {
      "epoch": 5.089269773254776,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 21378
    },
    {
      "epoch": 5.089507825983455,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 21379
    },
    {
      "epoch": 5.089745878712135,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 21380
    },
    {
      "epoch": 5.089983931440814,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 21381
    },
    {
      "epoch": 5.090221984169493,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 21382
    },
    {
      "epoch": 5.090460036898173,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 21383
    },
    {
      "epoch": 5.090698089626852,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 21384
    },
    {
      "epoch": 5.090936142355532,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 21385
    },
    {
      "epoch": 5.0911741950842115,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 21386
    },
    {
      "epoch": 5.09141224781289,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 21387
    },
    {
      "epoch": 5.09165030054157,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 21388
    },
    {
      "epoch": 5.0918883532702495,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 21389
    },
    {
      "epoch": 5.092126405998929,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 21390
    },
    {
      "epoch": 5.092364458727608,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 21391
    },
    {
      "epoch": 5.092602511456287,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 21392
    },
    {
      "epoch": 5.092840564184967,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9153,
      "step": 21393
    },
    {
      "epoch": 5.093078616913647,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 21394
    },
    {
      "epoch": 5.093316669642326,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 21395
    },
    {
      "epoch": 5.093554722371005,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4255,
      "step": 21396
    },
    {
      "epoch": 5.0937927750996845,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 21397
    },
    {
      "epoch": 5.094030827828364,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 21398
    },
    {
      "epoch": 5.094268880557044,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 21399
    },
    {
      "epoch": 5.0945069332857225,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 21400
    },
    {
      "epoch": 5.0945069332857225,
      "eval_runtime": 0.5327,
      "eval_samples_per_second": 1.877,
      "eval_steps_per_second": 1.877,
      "step": 21400
    },
    {
      "epoch": 5.094744986014402,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 21401
    },
    {
      "epoch": 5.094983038743082,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 21402
    },
    {
      "epoch": 5.095221091471761,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 21403
    },
    {
      "epoch": 5.09545914420044,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3407,
      "step": 21404
    },
    {
      "epoch": 5.09569719692912,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 21405
    },
    {
      "epoch": 5.095935249657799,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3396,
      "step": 21406
    },
    {
      "epoch": 5.096173302386479,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.438,
      "step": 21407
    },
    {
      "epoch": 5.096411355115158,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 21408
    },
    {
      "epoch": 5.096649407843837,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 21409
    },
    {
      "epoch": 5.096887460572517,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 21410
    },
    {
      "epoch": 5.097125513301196,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 21411
    },
    {
      "epoch": 5.097363566029876,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 21412
    },
    {
      "epoch": 5.097601618758555,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2649,
      "step": 21413
    },
    {
      "epoch": 5.097839671487234,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 21414
    },
    {
      "epoch": 5.098077724215914,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 21415
    },
    {
      "epoch": 5.0983157769445935,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 21416
    },
    {
      "epoch": 5.098553829673272,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 21417
    },
    {
      "epoch": 5.098791882401952,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 21418
    },
    {
      "epoch": 5.099029935130631,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 21419
    },
    {
      "epoch": 5.099267987859311,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 21420
    },
    {
      "epoch": 5.099506040587991,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 21421
    },
    {
      "epoch": 5.099744093316669,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 21422
    },
    {
      "epoch": 5.099982146045349,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 21423
    },
    {
      "epoch": 5.1002201987740285,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 21424
    },
    {
      "epoch": 5.100458251502708,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 21425
    },
    {
      "epoch": 5.100696304231387,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 21426
    },
    {
      "epoch": 5.1009343569600665,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 21427
    },
    {
      "epoch": 5.101172409688746,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 21428
    },
    {
      "epoch": 5.101410462417426,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 21429
    },
    {
      "epoch": 5.101648515146105,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 21430
    },
    {
      "epoch": 5.101886567874784,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 21431
    },
    {
      "epoch": 5.102124620603464,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 21432
    },
    {
      "epoch": 5.102362673332143,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 21433
    },
    {
      "epoch": 5.102600726060823,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 21434
    },
    {
      "epoch": 5.1028387787895015,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 21435
    },
    {
      "epoch": 5.103076831518181,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 21436
    },
    {
      "epoch": 5.103314884246861,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 21437
    },
    {
      "epoch": 5.10355293697554,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 21438
    },
    {
      "epoch": 5.103790989704219,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 21439
    },
    {
      "epoch": 5.104029042432899,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 21440
    },
    {
      "epoch": 5.104267095161578,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.977,
      "step": 21441
    },
    {
      "epoch": 5.104505147890258,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 21442
    },
    {
      "epoch": 5.1047432006189375,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 21443
    },
    {
      "epoch": 5.104981253347616,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 21444
    },
    {
      "epoch": 5.105219306076296,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 21445
    },
    {
      "epoch": 5.105457358804975,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 21446
    },
    {
      "epoch": 5.105695411533655,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 21447
    },
    {
      "epoch": 5.105933464262334,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 21448
    },
    {
      "epoch": 5.106171516991013,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 21449
    },
    {
      "epoch": 5.106409569719693,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 21450
    },
    {
      "epoch": 5.106647622448373,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 21451
    },
    {
      "epoch": 5.106885675177052,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 21452
    },
    {
      "epoch": 5.107123727905731,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 21453
    },
    {
      "epoch": 5.1073617806344105,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 21454
    },
    {
      "epoch": 5.10759983336309,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 21455
    },
    {
      "epoch": 5.10783788609177,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 21456
    },
    {
      "epoch": 5.108075938820448,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 21457
    },
    {
      "epoch": 5.108313991549128,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8643,
      "step": 21458
    },
    {
      "epoch": 5.108552044277808,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 21459
    },
    {
      "epoch": 5.108790097006487,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 21460
    },
    {
      "epoch": 5.109028149735166,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 21461
    },
    {
      "epoch": 5.1092662024638456,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 21462
    },
    {
      "epoch": 5.109504255192525,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 21463
    },
    {
      "epoch": 5.109742307921205,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 21464
    },
    {
      "epoch": 5.109980360649884,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 21465
    },
    {
      "epoch": 5.110218413378563,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 21466
    },
    {
      "epoch": 5.110456466107243,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 21467
    },
    {
      "epoch": 5.110694518835922,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 21468
    },
    {
      "epoch": 5.110932571564602,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4199,
      "step": 21469
    },
    {
      "epoch": 5.111170624293281,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 21470
    },
    {
      "epoch": 5.11140867702196,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 21471
    },
    {
      "epoch": 5.11164672975064,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 21472
    },
    {
      "epoch": 5.111884782479319,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 21473
    },
    {
      "epoch": 5.112122835207998,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 21474
    },
    {
      "epoch": 5.112360887936678,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 21475
    },
    {
      "epoch": 5.112598940665357,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 21476
    },
    {
      "epoch": 5.112836993394037,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 21477
    },
    {
      "epoch": 5.113075046122717,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 21478
    },
    {
      "epoch": 5.113313098851395,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 21479
    },
    {
      "epoch": 5.113551151580075,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 21480
    },
    {
      "epoch": 5.1137892043087545,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 21481
    },
    {
      "epoch": 5.114027257037434,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 21482
    },
    {
      "epoch": 5.114265309766113,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.5391,
      "step": 21483
    },
    {
      "epoch": 5.114503362494792,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 21484
    },
    {
      "epoch": 5.114741415223472,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 21485
    },
    {
      "epoch": 5.114979467952152,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3192,
      "step": 21486
    },
    {
      "epoch": 5.115217520680831,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 21487
    },
    {
      "epoch": 5.11545557340951,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 21488
    },
    {
      "epoch": 5.11569362613819,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 21489
    },
    {
      "epoch": 5.115931678866869,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 21490
    },
    {
      "epoch": 5.116169731595549,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 21491
    },
    {
      "epoch": 5.1164077843242275,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 21492
    },
    {
      "epoch": 5.116645837052907,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 21493
    },
    {
      "epoch": 5.116883889781587,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 21494
    },
    {
      "epoch": 5.117121942510266,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 21495
    },
    {
      "epoch": 5.117359995238945,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 21496
    },
    {
      "epoch": 5.117598047967625,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 21497
    },
    {
      "epoch": 5.117836100696304,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 21498
    },
    {
      "epoch": 5.118074153424984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 21499
    },
    {
      "epoch": 5.1183122061536634,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 21500
    },
    {
      "epoch": 5.1183122061536634,
      "eval_runtime": 0.5258,
      "eval_samples_per_second": 1.902,
      "eval_steps_per_second": 1.902,
      "step": 21500
    },
    {
      "epoch": 5.118550258882342,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 21501
    },
    {
      "epoch": 5.118788311611022,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 21502
    },
    {
      "epoch": 5.119026364339701,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 21503
    },
    {
      "epoch": 5.119264417068381,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 21504
    },
    {
      "epoch": 5.11950246979706,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 21505
    },
    {
      "epoch": 5.119740522525739,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 21506
    },
    {
      "epoch": 5.119978575254419,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 21507
    },
    {
      "epoch": 5.1202166279830985,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 21508
    },
    {
      "epoch": 5.120454680711777,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4232,
      "step": 21509
    },
    {
      "epoch": 5.120692733440457,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 21510
    },
    {
      "epoch": 5.120930786169136,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 21511
    },
    {
      "epoch": 5.121168838897816,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 21512
    },
    {
      "epoch": 5.121406891626496,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 21513
    },
    {
      "epoch": 5.121644944355174,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 21514
    },
    {
      "epoch": 5.121882997083854,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8626,
      "step": 21515
    },
    {
      "epoch": 5.122121049812534,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 21516
    },
    {
      "epoch": 5.122359102541213,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 21517
    },
    {
      "epoch": 5.122597155269892,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.5862,
      "step": 21518
    },
    {
      "epoch": 5.1228352079985715,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 21519
    },
    {
      "epoch": 5.123073260727251,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 21520
    },
    {
      "epoch": 5.123311313455931,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 21521
    },
    {
      "epoch": 5.12354936618461,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 21522
    },
    {
      "epoch": 5.123787418913289,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 21523
    },
    {
      "epoch": 5.124025471641969,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 21524
    },
    {
      "epoch": 5.124263524370648,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 21525
    },
    {
      "epoch": 5.124501577099328,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 21526
    },
    {
      "epoch": 5.124739629828007,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 21527
    },
    {
      "epoch": 5.124977682556686,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 21528
    },
    {
      "epoch": 5.125215735285366,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 21529
    },
    {
      "epoch": 5.125453788014045,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 21530
    },
    {
      "epoch": 5.125691840742724,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 21531
    },
    {
      "epoch": 5.125929893471404,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 21532
    },
    {
      "epoch": 5.126167946200083,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 21533
    },
    {
      "epoch": 5.126405998928763,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 21534
    },
    {
      "epoch": 5.1266440516574425,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 21535
    },
    {
      "epoch": 5.126882104386121,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 21536
    },
    {
      "epoch": 5.127120157114801,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 21537
    },
    {
      "epoch": 5.1273582098434805,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 21538
    },
    {
      "epoch": 5.12759626257216,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 21539
    },
    {
      "epoch": 5.127834315300839,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 21540
    },
    {
      "epoch": 5.128072368029518,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 21541
    },
    {
      "epoch": 5.128310420758198,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 21542
    },
    {
      "epoch": 5.128548473486878,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 21543
    },
    {
      "epoch": 5.128786526215556,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 21544
    },
    {
      "epoch": 5.129024578944236,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 21545
    },
    {
      "epoch": 5.1292626316729155,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 21546
    },
    {
      "epoch": 5.129500684401595,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 21547
    },
    {
      "epoch": 5.129738737130275,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 21548
    },
    {
      "epoch": 5.1299767898589534,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 21549
    },
    {
      "epoch": 5.130214842587633,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 21550
    },
    {
      "epoch": 5.130452895316313,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 21551
    },
    {
      "epoch": 5.130690948044992,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 21552
    },
    {
      "epoch": 5.130929000773671,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 21553
    },
    {
      "epoch": 5.131167053502351,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8649,
      "step": 21554
    },
    {
      "epoch": 5.13140510623103,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 21555
    },
    {
      "epoch": 5.13164315895971,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.785,
      "step": 21556
    },
    {
      "epoch": 5.131881211688389,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 21557
    },
    {
      "epoch": 5.132119264417068,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 21558
    },
    {
      "epoch": 5.132357317145748,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 21559
    },
    {
      "epoch": 5.132595369874427,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 21560
    },
    {
      "epoch": 5.132833422603107,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 21561
    },
    {
      "epoch": 5.133071475331786,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 21562
    },
    {
      "epoch": 5.133309528060465,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 21563
    },
    {
      "epoch": 5.133547580789145,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 21564
    },
    {
      "epoch": 5.1337856335178245,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 21565
    },
    {
      "epoch": 5.134023686246503,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 21566
    },
    {
      "epoch": 5.134261738975183,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 21567
    },
    {
      "epoch": 5.134499791703862,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 21568
    },
    {
      "epoch": 5.134737844432542,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 21569
    },
    {
      "epoch": 5.134975897161222,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 21570
    },
    {
      "epoch": 5.1352139498899,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 21571
    },
    {
      "epoch": 5.13545200261858,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 21572
    },
    {
      "epoch": 5.1356900553472595,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 21573
    },
    {
      "epoch": 5.135928108075939,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 21574
    },
    {
      "epoch": 5.136166160804618,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 21575
    },
    {
      "epoch": 5.1364042135332975,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 21576
    },
    {
      "epoch": 5.136642266261977,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 21577
    },
    {
      "epoch": 5.136880318990657,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 21578
    },
    {
      "epoch": 5.137118371719335,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 21579
    },
    {
      "epoch": 5.137356424448015,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 21580
    },
    {
      "epoch": 5.137594477176695,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 21581
    },
    {
      "epoch": 5.137832529905374,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8541,
      "step": 21582
    },
    {
      "epoch": 5.138070582634054,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 21583
    },
    {
      "epoch": 5.1383086353627325,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3293,
      "step": 21584
    },
    {
      "epoch": 5.138546688091412,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 21585
    },
    {
      "epoch": 5.138784740820092,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 21586
    },
    {
      "epoch": 5.139022793548771,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 21587
    },
    {
      "epoch": 5.13926084627745,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 21588
    },
    {
      "epoch": 5.13949889900613,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 21589
    },
    {
      "epoch": 5.139736951734809,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 21590
    },
    {
      "epoch": 5.139975004463489,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 21591
    },
    {
      "epoch": 5.1402130571921685,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 21592
    },
    {
      "epoch": 5.140451109920847,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 21593
    },
    {
      "epoch": 5.140689162649527,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 21594
    },
    {
      "epoch": 5.140927215378206,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 21595
    },
    {
      "epoch": 5.141165268106886,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 21596
    },
    {
      "epoch": 5.141403320835565,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3927,
      "step": 21597
    },
    {
      "epoch": 5.141641373564244,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 21598
    },
    {
      "epoch": 5.141879426292924,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 21599
    },
    {
      "epoch": 5.1421174790216035,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 21600
    },
    {
      "epoch": 5.1421174790216035,
      "eval_runtime": 0.5291,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 21600
    },
    {
      "epoch": 5.142355531750282,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 21601
    },
    {
      "epoch": 5.142593584478962,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 21602
    },
    {
      "epoch": 5.1428316372076415,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 21603
    },
    {
      "epoch": 5.143069689936321,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 21604
    },
    {
      "epoch": 5.143307742665001,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 21605
    },
    {
      "epoch": 5.143545795393679,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 21606
    },
    {
      "epoch": 5.143783848122359,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2772,
      "step": 21607
    },
    {
      "epoch": 5.144021900851039,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 21608
    },
    {
      "epoch": 5.144259953579718,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 21609
    },
    {
      "epoch": 5.144498006308397,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 21610
    },
    {
      "epoch": 5.1447360590370765,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 21611
    },
    {
      "epoch": 5.144974111765756,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 21612
    },
    {
      "epoch": 5.145212164494436,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 21613
    },
    {
      "epoch": 5.1454502172231145,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 21614
    },
    {
      "epoch": 5.145688269951794,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 21615
    },
    {
      "epoch": 5.145926322680474,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 21616
    },
    {
      "epoch": 5.146164375409153,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.3627,
      "step": 21617
    },
    {
      "epoch": 5.146402428137833,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 21618
    },
    {
      "epoch": 5.146640480866512,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 21619
    },
    {
      "epoch": 5.146878533595191,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4452,
      "step": 21620
    },
    {
      "epoch": 5.147116586323871,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 21621
    },
    {
      "epoch": 5.14735463905255,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 21622
    },
    {
      "epoch": 5.147592691781229,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 21623
    },
    {
      "epoch": 5.147830744509909,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 21624
    },
    {
      "epoch": 5.148068797238588,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 21625
    },
    {
      "epoch": 5.148306849967268,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 21626
    },
    {
      "epoch": 5.148544902695948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 21627
    },
    {
      "epoch": 5.148782955424626,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 21628
    },
    {
      "epoch": 5.149021008153306,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 21629
    },
    {
      "epoch": 5.1492590608819855,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 21630
    },
    {
      "epoch": 5.149497113610665,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 21631
    },
    {
      "epoch": 5.149735166339344,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 21632
    },
    {
      "epoch": 5.149973219068023,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 21633
    },
    {
      "epoch": 5.150211271796703,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3617,
      "step": 21634
    },
    {
      "epoch": 5.150449324525383,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 21635
    },
    {
      "epoch": 5.150687377254062,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 21636
    },
    {
      "epoch": 5.150925429982741,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 21637
    },
    {
      "epoch": 5.1511634827114205,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 21638
    },
    {
      "epoch": 5.1514015354401,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 21639
    },
    {
      "epoch": 5.15163958816878,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 21640
    },
    {
      "epoch": 5.1518776408974585,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 21641
    },
    {
      "epoch": 5.152115693626138,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 21642
    },
    {
      "epoch": 5.152353746354818,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 21643
    },
    {
      "epoch": 5.152591799083497,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 21644
    },
    {
      "epoch": 5.152829851812176,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.5421,
      "step": 21645
    },
    {
      "epoch": 5.153067904540856,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 21646
    },
    {
      "epoch": 5.153305957269535,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 21647
    },
    {
      "epoch": 5.153544009998215,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 21648
    },
    {
      "epoch": 5.153782062726894,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 21649
    },
    {
      "epoch": 5.154020115455573,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 21650
    },
    {
      "epoch": 5.154258168184253,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 21651
    },
    {
      "epoch": 5.154496220912932,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 21652
    },
    {
      "epoch": 5.154734273641612,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5534,
      "step": 21653
    },
    {
      "epoch": 5.154972326370291,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 21654
    },
    {
      "epoch": 5.15521037909897,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 21655
    },
    {
      "epoch": 5.15544843182765,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 21656
    },
    {
      "epoch": 5.1556864845563295,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 21657
    },
    {
      "epoch": 5.155924537285008,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 21658
    },
    {
      "epoch": 5.156162590013688,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.5543,
      "step": 21659
    },
    {
      "epoch": 5.156400642742367,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 21660
    },
    {
      "epoch": 5.156638695471047,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 21661
    },
    {
      "epoch": 5.156876748199727,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 21662
    },
    {
      "epoch": 5.157114800928405,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 21663
    },
    {
      "epoch": 5.157352853657085,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 21664
    },
    {
      "epoch": 5.157590906385765,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5193,
      "step": 21665
    },
    {
      "epoch": 5.157828959114444,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 21666
    },
    {
      "epoch": 5.158067011843123,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 21667
    },
    {
      "epoch": 5.1583050645718025,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 21668
    },
    {
      "epoch": 5.158543117300482,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 21669
    },
    {
      "epoch": 5.158781170029162,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 21670
    },
    {
      "epoch": 5.159019222757841,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 21671
    },
    {
      "epoch": 5.15925727548652,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 21672
    },
    {
      "epoch": 5.1594953282152,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 21673
    },
    {
      "epoch": 5.159733380943879,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 21674
    },
    {
      "epoch": 5.159971433672559,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 21675
    },
    {
      "epoch": 5.1602094864012376,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 21676
    },
    {
      "epoch": 5.160447539129917,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 21677
    },
    {
      "epoch": 5.160685591858597,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 21678
    },
    {
      "epoch": 5.160923644587276,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 21679
    },
    {
      "epoch": 5.161161697315955,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 21680
    },
    {
      "epoch": 5.161399750044635,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 21681
    },
    {
      "epoch": 5.161637802773314,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 21682
    },
    {
      "epoch": 5.161875855501994,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 21683
    },
    {
      "epoch": 5.1621139082306735,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 21684
    },
    {
      "epoch": 5.162351960959352,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 21685
    },
    {
      "epoch": 5.162590013688032,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2894,
      "step": 21686
    },
    {
      "epoch": 5.162828066416711,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 21687
    },
    {
      "epoch": 5.163066119145391,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 21688
    },
    {
      "epoch": 5.16330417187407,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 21689
    },
    {
      "epoch": 5.163542224602749,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 21690
    },
    {
      "epoch": 5.163780277331429,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 21691
    },
    {
      "epoch": 5.164018330060109,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3017,
      "step": 21692
    },
    {
      "epoch": 5.164256382788787,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 21693
    },
    {
      "epoch": 5.164494435517467,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 21694
    },
    {
      "epoch": 5.1647324882461465,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 21695
    },
    {
      "epoch": 5.164970540974826,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 21696
    },
    {
      "epoch": 5.165208593703506,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7338,
      "step": 21697
    },
    {
      "epoch": 5.165446646432184,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 21698
    },
    {
      "epoch": 5.165684699160864,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 21699
    },
    {
      "epoch": 5.165922751889544,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 21700
    },
    {
      "epoch": 5.165922751889544,
      "eval_runtime": 0.5208,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 21700
    },
    {
      "epoch": 5.166160804618223,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 21701
    },
    {
      "epoch": 5.166398857346902,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 21702
    },
    {
      "epoch": 5.166636910075582,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 21703
    },
    {
      "epoch": 5.166874962804261,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 21704
    },
    {
      "epoch": 5.167113015532941,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 21705
    },
    {
      "epoch": 5.16735106826162,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 21706
    },
    {
      "epoch": 5.167589120990299,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 21707
    },
    {
      "epoch": 5.167827173718979,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 21708
    },
    {
      "epoch": 5.168065226447658,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 21709
    },
    {
      "epoch": 5.168303279176338,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 21710
    },
    {
      "epoch": 5.168541331905017,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 21711
    },
    {
      "epoch": 5.168779384633696,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 21712
    },
    {
      "epoch": 5.169017437362376,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 21713
    },
    {
      "epoch": 5.1692554900910554,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 21714
    },
    {
      "epoch": 5.169493542819734,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 21715
    },
    {
      "epoch": 5.169731595548414,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 21716
    },
    {
      "epoch": 5.169969648277093,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 21717
    },
    {
      "epoch": 5.170207701005773,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 21718
    },
    {
      "epoch": 5.170445753734453,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 21719
    },
    {
      "epoch": 5.170683806463131,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 21720
    },
    {
      "epoch": 5.170921859191811,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 21721
    },
    {
      "epoch": 5.1711599119204905,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 21722
    },
    {
      "epoch": 5.17139796464917,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 21723
    },
    {
      "epoch": 5.171636017377849,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 21724
    },
    {
      "epoch": 5.171874070106528,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 21725
    },
    {
      "epoch": 5.172112122835208,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 21726
    },
    {
      "epoch": 5.172350175563888,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 21727
    },
    {
      "epoch": 5.172588228292566,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 21728
    },
    {
      "epoch": 5.172826281021246,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 21729
    },
    {
      "epoch": 5.173064333749926,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 21730
    },
    {
      "epoch": 5.173302386478605,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 21731
    },
    {
      "epoch": 5.173540439207285,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 21732
    },
    {
      "epoch": 5.1737784919359635,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 21733
    },
    {
      "epoch": 5.174016544664643,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 21734
    },
    {
      "epoch": 5.174254597393323,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 21735
    },
    {
      "epoch": 5.174492650122002,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 21736
    },
    {
      "epoch": 5.174730702850681,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 21737
    },
    {
      "epoch": 5.174968755579361,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 21738
    },
    {
      "epoch": 5.17520680830804,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.422,
      "step": 21739
    },
    {
      "epoch": 5.17544486103672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 21740
    },
    {
      "epoch": 5.1756829137653995,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 21741
    },
    {
      "epoch": 5.175920966494078,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 21742
    },
    {
      "epoch": 5.176159019222758,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 21743
    },
    {
      "epoch": 5.176397071951437,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 21744
    },
    {
      "epoch": 5.176635124680117,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 21745
    },
    {
      "epoch": 5.176873177408796,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 21746
    },
    {
      "epoch": 5.177111230137475,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 21747
    },
    {
      "epoch": 5.177349282866155,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 21748
    },
    {
      "epoch": 5.1775873355948345,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 21749
    },
    {
      "epoch": 5.177825388323513,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 21750
    },
    {
      "epoch": 5.178063441052193,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 21751
    },
    {
      "epoch": 5.1783014937808725,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 21752
    },
    {
      "epoch": 5.178539546509552,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 21753
    },
    {
      "epoch": 5.178777599238232,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 21754
    },
    {
      "epoch": 5.17901565196691,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 21755
    },
    {
      "epoch": 5.17925370469559,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 21756
    },
    {
      "epoch": 5.17949175742427,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 21757
    },
    {
      "epoch": 5.179729810152949,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 21758
    },
    {
      "epoch": 5.179967862881628,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 21759
    },
    {
      "epoch": 5.1802059156103075,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 21760
    },
    {
      "epoch": 5.180443968338987,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 21761
    },
    {
      "epoch": 5.180682021067667,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 21762
    },
    {
      "epoch": 5.1809200737963454,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3871,
      "step": 21763
    },
    {
      "epoch": 5.181158126525025,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 21764
    },
    {
      "epoch": 5.181396179253705,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 21765
    },
    {
      "epoch": 5.181634231982384,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 21766
    },
    {
      "epoch": 5.181872284711064,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 21767
    },
    {
      "epoch": 5.182110337439743,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 21768
    },
    {
      "epoch": 5.182348390168422,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 21769
    },
    {
      "epoch": 5.182586442897102,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2244,
      "step": 21770
    },
    {
      "epoch": 5.182824495625781,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 21771
    },
    {
      "epoch": 5.18306254835446,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 21772
    },
    {
      "epoch": 5.18330060108314,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 21773
    },
    {
      "epoch": 5.183538653811819,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 21774
    },
    {
      "epoch": 5.183776706540499,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 21775
    },
    {
      "epoch": 5.1840147592691785,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 21776
    },
    {
      "epoch": 5.184252811997857,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 21777
    },
    {
      "epoch": 5.184490864726537,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2823,
      "step": 21778
    },
    {
      "epoch": 5.1847289174552165,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 21779
    },
    {
      "epoch": 5.184966970183896,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 21780
    },
    {
      "epoch": 5.185205022912575,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 21781
    },
    {
      "epoch": 5.185443075641254,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 21782
    },
    {
      "epoch": 5.185681128369934,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 21783
    },
    {
      "epoch": 5.185919181098614,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 21784
    },
    {
      "epoch": 5.186157233827292,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 21785
    },
    {
      "epoch": 5.186395286555972,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 21786
    },
    {
      "epoch": 5.1866333392846515,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 21787
    },
    {
      "epoch": 5.186871392013331,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3536,
      "step": 21788
    },
    {
      "epoch": 5.187109444742011,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 21789
    },
    {
      "epoch": 5.1873474974706895,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 21790
    },
    {
      "epoch": 5.187585550199369,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 21791
    },
    {
      "epoch": 5.187823602928049,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 21792
    },
    {
      "epoch": 5.188061655656728,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 21793
    },
    {
      "epoch": 5.188299708385407,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 21794
    },
    {
      "epoch": 5.188537761114087,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 21795
    },
    {
      "epoch": 5.188775813842766,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 21796
    },
    {
      "epoch": 5.189013866571446,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 21797
    },
    {
      "epoch": 5.1892519193001245,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 21798
    },
    {
      "epoch": 5.189489972028804,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 21799
    },
    {
      "epoch": 5.189728024757484,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 21800
    },
    {
      "epoch": 5.189728024757484,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 21800
    },
    {
      "epoch": 5.189966077486163,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 21801
    },
    {
      "epoch": 5.190204130214843,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 21802
    },
    {
      "epoch": 5.190442182943522,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 21803
    },
    {
      "epoch": 5.190680235672201,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 21804
    },
    {
      "epoch": 5.190918288400881,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 21805
    },
    {
      "epoch": 5.1911563411295605,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3073,
      "step": 21806
    },
    {
      "epoch": 5.191394393858239,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 21807
    },
    {
      "epoch": 5.191632446586919,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 21808
    },
    {
      "epoch": 5.191870499315598,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 21809
    },
    {
      "epoch": 5.192108552044278,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 21810
    },
    {
      "epoch": 5.192346604772958,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 21811
    },
    {
      "epoch": 5.192584657501636,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 21812
    },
    {
      "epoch": 5.192822710230316,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 21813
    },
    {
      "epoch": 5.1930607629589955,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 21814
    },
    {
      "epoch": 5.193298815687675,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 21815
    },
    {
      "epoch": 5.193536868416354,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 21816
    },
    {
      "epoch": 5.1937749211450335,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 21817
    },
    {
      "epoch": 5.194012973873713,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 21818
    },
    {
      "epoch": 5.194251026602393,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 21819
    },
    {
      "epoch": 5.194489079331072,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 21820
    },
    {
      "epoch": 5.194727132059751,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2343,
      "step": 21821
    },
    {
      "epoch": 5.194965184788431,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 21822
    },
    {
      "epoch": 5.19520323751711,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 21823
    },
    {
      "epoch": 5.19544129024579,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 21824
    },
    {
      "epoch": 5.1956793429744685,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3286,
      "step": 21825
    },
    {
      "epoch": 5.195917395703148,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 21826
    },
    {
      "epoch": 5.196155448431828,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1647,
      "step": 21827
    },
    {
      "epoch": 5.196393501160507,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 21828
    },
    {
      "epoch": 5.196631553889186,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 21829
    },
    {
      "epoch": 5.196869606617866,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 21830
    },
    {
      "epoch": 5.197107659346545,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4401,
      "step": 21831
    },
    {
      "epoch": 5.197345712075225,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 21832
    },
    {
      "epoch": 5.1975837648039045,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 21833
    },
    {
      "epoch": 5.197821817532583,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 21834
    },
    {
      "epoch": 5.198059870261263,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 21835
    },
    {
      "epoch": 5.198297922989942,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 21836
    },
    {
      "epoch": 5.198535975718622,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 21837
    },
    {
      "epoch": 5.198774028447301,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3613,
      "step": 21838
    },
    {
      "epoch": 5.19901208117598,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 21839
    },
    {
      "epoch": 5.19925013390466,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 21840
    },
    {
      "epoch": 5.19948818663334,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 21841
    },
    {
      "epoch": 5.199726239362018,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 21842
    },
    {
      "epoch": 5.199964292090698,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 21843
    },
    {
      "epoch": 5.2002023448193775,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 21844
    },
    {
      "epoch": 5.200440397548057,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 21845
    },
    {
      "epoch": 5.200678450276737,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 21846
    },
    {
      "epoch": 5.200916503005415,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 21847
    },
    {
      "epoch": 5.201154555734095,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 21848
    },
    {
      "epoch": 5.201392608462775,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 21849
    },
    {
      "epoch": 5.201630661191454,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 21850
    },
    {
      "epoch": 5.201868713920133,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 21851
    },
    {
      "epoch": 5.2021067666488126,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 21852
    },
    {
      "epoch": 5.202344819377492,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 21853
    },
    {
      "epoch": 5.202582872106172,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.989,
      "step": 21854
    },
    {
      "epoch": 5.202820924834851,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 21855
    },
    {
      "epoch": 5.20305897756353,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 21856
    },
    {
      "epoch": 5.20329703029221,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 21857
    },
    {
      "epoch": 5.203535083020889,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 21858
    },
    {
      "epoch": 5.203773135749569,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 21859
    },
    {
      "epoch": 5.204011188478248,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 21860
    },
    {
      "epoch": 5.204249241206927,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 21861
    },
    {
      "epoch": 5.204487293935607,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 21862
    },
    {
      "epoch": 5.204725346664286,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 21863
    },
    {
      "epoch": 5.204963399392965,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8411,
      "step": 21864
    },
    {
      "epoch": 5.205201452121645,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 21865
    },
    {
      "epoch": 5.205439504850324,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 21866
    },
    {
      "epoch": 5.205677557579004,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 21867
    },
    {
      "epoch": 5.205915610307684,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 21868
    },
    {
      "epoch": 5.206153663036362,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 21869
    },
    {
      "epoch": 5.206391715765042,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 21870
    },
    {
      "epoch": 5.2066297684937215,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 21871
    },
    {
      "epoch": 5.206867821222401,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 21872
    },
    {
      "epoch": 5.20710587395108,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 21873
    },
    {
      "epoch": 5.207343926679759,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 21874
    },
    {
      "epoch": 5.207581979408439,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2825,
      "step": 21875
    },
    {
      "epoch": 5.207820032137119,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 21876
    },
    {
      "epoch": 5.208058084865797,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 21877
    },
    {
      "epoch": 5.208296137594477,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 21878
    },
    {
      "epoch": 5.208534190323157,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2367,
      "step": 21879
    },
    {
      "epoch": 5.208772243051836,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 21880
    },
    {
      "epoch": 5.209010295780516,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 21881
    },
    {
      "epoch": 5.2092483485091945,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 21882
    },
    {
      "epoch": 5.209486401237874,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 21883
    },
    {
      "epoch": 5.209724453966554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 21884
    },
    {
      "epoch": 5.209962506695233,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 21885
    },
    {
      "epoch": 5.210200559423912,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 21886
    },
    {
      "epoch": 5.210438612152592,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 21887
    },
    {
      "epoch": 5.210676664881271,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 21888
    },
    {
      "epoch": 5.210914717609951,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 21889
    },
    {
      "epoch": 5.2111527703386304,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 21890
    },
    {
      "epoch": 5.211390823067309,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 21891
    },
    {
      "epoch": 5.211628875795989,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 21892
    },
    {
      "epoch": 5.211866928524668,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 21893
    },
    {
      "epoch": 5.212104981253348,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 21894
    },
    {
      "epoch": 5.212343033982027,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 21895
    },
    {
      "epoch": 5.212581086710706,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7916,
      "step": 21896
    },
    {
      "epoch": 5.212819139439386,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 21897
    },
    {
      "epoch": 5.2130571921680655,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 21898
    },
    {
      "epoch": 5.213295244896744,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 21899
    },
    {
      "epoch": 5.213533297625424,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 21900
    },
    {
      "epoch": 5.213533297625424,
      "eval_runtime": 0.5266,
      "eval_samples_per_second": 1.899,
      "eval_steps_per_second": 1.899,
      "step": 21900
    },
    {
      "epoch": 5.213771350354103,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 21901
    },
    {
      "epoch": 5.214009403082783,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 21902
    },
    {
      "epoch": 5.214247455811463,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3435,
      "step": 21903
    },
    {
      "epoch": 5.214485508540141,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9867,
      "step": 21904
    },
    {
      "epoch": 5.214723561268821,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 21905
    },
    {
      "epoch": 5.214961613997501,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 21906
    },
    {
      "epoch": 5.21519966672618,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 21907
    },
    {
      "epoch": 5.215437719454859,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 21908
    },
    {
      "epoch": 5.2156757721835385,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 21909
    },
    {
      "epoch": 5.215913824912218,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 21910
    },
    {
      "epoch": 5.216151877640898,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 21911
    },
    {
      "epoch": 5.216389930369576,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 21912
    },
    {
      "epoch": 5.216627983098256,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 21913
    },
    {
      "epoch": 5.216866035826936,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 21914
    },
    {
      "epoch": 5.217104088555615,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 21915
    },
    {
      "epoch": 5.217342141284295,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 21916
    },
    {
      "epoch": 5.217580194012974,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 21917
    },
    {
      "epoch": 5.217818246741653,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 21918
    },
    {
      "epoch": 5.218056299470333,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 21919
    },
    {
      "epoch": 5.218294352199012,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4681,
      "step": 21920
    },
    {
      "epoch": 5.218532404927691,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 21921
    },
    {
      "epoch": 5.218770457656371,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 21922
    },
    {
      "epoch": 5.21900851038505,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 21923
    },
    {
      "epoch": 5.21924656311373,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 21924
    },
    {
      "epoch": 5.2194846158424095,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5922,
      "step": 21925
    },
    {
      "epoch": 5.219722668571088,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 21926
    },
    {
      "epoch": 5.219960721299768,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 21927
    },
    {
      "epoch": 5.2201987740284475,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5592,
      "step": 21928
    },
    {
      "epoch": 5.220436826757127,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 21929
    },
    {
      "epoch": 5.220674879485806,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 21930
    },
    {
      "epoch": 5.220912932214485,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 21931
    },
    {
      "epoch": 5.221150984943165,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 21932
    },
    {
      "epoch": 5.221389037671845,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 21933
    },
    {
      "epoch": 5.221627090400523,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 21934
    },
    {
      "epoch": 5.221865143129203,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 21935
    },
    {
      "epoch": 5.2221031958578825,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 21936
    },
    {
      "epoch": 5.222341248586562,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 21937
    },
    {
      "epoch": 5.222579301315242,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 21938
    },
    {
      "epoch": 5.2228173540439204,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 21939
    },
    {
      "epoch": 5.2230554067726,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 21940
    },
    {
      "epoch": 5.22329345950128,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 21941
    },
    {
      "epoch": 5.223531512229959,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 21942
    },
    {
      "epoch": 5.223769564958638,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 21943
    },
    {
      "epoch": 5.224007617687318,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 21944
    },
    {
      "epoch": 5.224245670415997,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 21945
    },
    {
      "epoch": 5.224483723144677,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 21946
    },
    {
      "epoch": 5.2247217758733555,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 21947
    },
    {
      "epoch": 5.224959828602035,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 21948
    },
    {
      "epoch": 5.225197881330715,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 21949
    },
    {
      "epoch": 5.225435934059394,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 21950
    },
    {
      "epoch": 5.225673986788074,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 21951
    },
    {
      "epoch": 5.225912039516753,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 21952
    },
    {
      "epoch": 5.226150092245432,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 21953
    },
    {
      "epoch": 5.226388144974112,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 21954
    },
    {
      "epoch": 5.2266261977027915,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 21955
    },
    {
      "epoch": 5.22686425043147,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4408,
      "step": 21956
    },
    {
      "epoch": 5.22710230316015,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 21957
    },
    {
      "epoch": 5.227340355888829,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 21958
    },
    {
      "epoch": 5.227578408617509,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 21959
    },
    {
      "epoch": 5.227816461346189,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 21960
    },
    {
      "epoch": 5.228054514074867,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 21961
    },
    {
      "epoch": 5.228292566803547,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 21962
    },
    {
      "epoch": 5.2285306195322265,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 21963
    },
    {
      "epoch": 5.228768672260906,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 21964
    },
    {
      "epoch": 5.229006724989585,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 21965
    },
    {
      "epoch": 5.2292447777182645,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 21966
    },
    {
      "epoch": 5.229482830446944,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 21967
    },
    {
      "epoch": 5.229720883175624,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 21968
    },
    {
      "epoch": 5.229958935904302,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 21969
    },
    {
      "epoch": 5.230196988632982,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 21970
    },
    {
      "epoch": 5.230435041361662,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 21971
    },
    {
      "epoch": 5.230673094090341,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2892,
      "step": 21972
    },
    {
      "epoch": 5.230911146819021,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 21973
    },
    {
      "epoch": 5.2311491995476995,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 21974
    },
    {
      "epoch": 5.231387252276379,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.5379,
      "step": 21975
    },
    {
      "epoch": 5.231625305005059,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3232,
      "step": 21976
    },
    {
      "epoch": 5.231863357733738,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 21977
    },
    {
      "epoch": 5.232101410462417,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 21978
    },
    {
      "epoch": 5.232339463191097,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.339,
      "step": 21979
    },
    {
      "epoch": 5.232577515919776,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 21980
    },
    {
      "epoch": 5.232815568648456,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 21981
    },
    {
      "epoch": 5.233053621377135,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 21982
    },
    {
      "epoch": 5.233291674105814,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 21983
    },
    {
      "epoch": 5.233529726834494,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 21984
    },
    {
      "epoch": 5.233767779563173,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 21985
    },
    {
      "epoch": 5.234005832291853,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 21986
    },
    {
      "epoch": 5.234243885020532,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 21987
    },
    {
      "epoch": 5.234481937749211,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 21988
    },
    {
      "epoch": 5.234719990477891,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 21989
    },
    {
      "epoch": 5.2349580432065705,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 21990
    },
    {
      "epoch": 5.235196095935249,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 21991
    },
    {
      "epoch": 5.235434148663929,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 21992
    },
    {
      "epoch": 5.2356722013926085,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 21993
    },
    {
      "epoch": 5.235910254121288,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 21994
    },
    {
      "epoch": 5.236148306849968,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 21995
    },
    {
      "epoch": 5.236386359578646,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1017,
      "step": 21996
    },
    {
      "epoch": 5.236624412307326,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 21997
    },
    {
      "epoch": 5.236862465036006,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 21998
    },
    {
      "epoch": 5.237100517764685,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 21999
    },
    {
      "epoch": 5.237338570493364,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 22000
    },
    {
      "epoch": 5.237338570493364,
      "eval_runtime": 0.5224,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 22000
    },
    {
      "epoch": 5.2375766232220435,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4041,
      "step": 22001
    },
    {
      "epoch": 5.237814675950723,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 22002
    },
    {
      "epoch": 5.238052728679403,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 22003
    },
    {
      "epoch": 5.2382907814080815,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 22004
    },
    {
      "epoch": 5.238528834136761,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 22005
    },
    {
      "epoch": 5.238766886865441,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 22006
    },
    {
      "epoch": 5.23900493959412,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 22007
    },
    {
      "epoch": 5.2392429923228,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 22008
    },
    {
      "epoch": 5.239481045051479,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 22009
    },
    {
      "epoch": 5.239719097780158,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 22010
    },
    {
      "epoch": 5.239957150508838,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 22011
    },
    {
      "epoch": 5.240195203237517,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 22012
    },
    {
      "epoch": 5.240433255966196,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 22013
    },
    {
      "epoch": 5.240671308694876,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 22014
    },
    {
      "epoch": 5.240909361423555,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 22015
    },
    {
      "epoch": 5.241147414152235,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 22016
    },
    {
      "epoch": 5.241385466880914,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 22017
    },
    {
      "epoch": 5.241623519609593,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7524,
      "step": 22018
    },
    {
      "epoch": 5.241861572338273,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4125,
      "step": 22019
    },
    {
      "epoch": 5.2420996250669525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 22020
    },
    {
      "epoch": 5.242337677795632,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 22021
    },
    {
      "epoch": 5.242575730524311,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 22022
    },
    {
      "epoch": 5.24281378325299,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 22023
    },
    {
      "epoch": 5.24305183598167,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 22024
    },
    {
      "epoch": 5.24328988871035,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 22025
    },
    {
      "epoch": 5.243527941439028,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 22026
    },
    {
      "epoch": 5.243765994167708,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 22027
    },
    {
      "epoch": 5.2440040468963875,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 22028
    },
    {
      "epoch": 5.244242099625067,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 22029
    },
    {
      "epoch": 5.244480152353747,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 22030
    },
    {
      "epoch": 5.2447182050824255,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 22031
    },
    {
      "epoch": 5.244956257811105,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 22032
    },
    {
      "epoch": 5.245194310539785,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 22033
    },
    {
      "epoch": 5.245432363268464,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 22034
    },
    {
      "epoch": 5.245670415997143,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 22035
    },
    {
      "epoch": 5.245908468725823,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 22036
    },
    {
      "epoch": 5.246146521454502,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 22037
    },
    {
      "epoch": 5.246384574183182,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2759,
      "step": 22038
    },
    {
      "epoch": 5.246622626911861,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 22039
    },
    {
      "epoch": 5.24686067964054,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.3183,
      "step": 22040
    },
    {
      "epoch": 5.24709873236922,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 22041
    },
    {
      "epoch": 5.247336785097899,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 22042
    },
    {
      "epoch": 5.247574837826579,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 22043
    },
    {
      "epoch": 5.247812890555258,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 22044
    },
    {
      "epoch": 5.248050943283937,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 22045
    },
    {
      "epoch": 5.248288996012617,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 22046
    },
    {
      "epoch": 5.2485270487412965,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 22047
    },
    {
      "epoch": 5.248765101469975,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 22048
    },
    {
      "epoch": 5.249003154198655,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 22049
    },
    {
      "epoch": 5.249241206927334,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1681,
      "step": 22050
    },
    {
      "epoch": 5.249479259656014,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 22051
    },
    {
      "epoch": 5.249717312384694,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 22052
    },
    {
      "epoch": 5.249955365113372,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 22053
    },
    {
      "epoch": 5.250193417842052,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3371,
      "step": 22054
    },
    {
      "epoch": 5.250431470570732,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 22055
    },
    {
      "epoch": 5.250669523299411,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 22056
    },
    {
      "epoch": 5.25090757602809,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 22057
    },
    {
      "epoch": 5.2511456287567695,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 22058
    },
    {
      "epoch": 5.251383681485449,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2848,
      "step": 22059
    },
    {
      "epoch": 5.251621734214129,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 22060
    },
    {
      "epoch": 5.251859786942807,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 22061
    },
    {
      "epoch": 5.252097839671487,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 22062
    },
    {
      "epoch": 5.252335892400167,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2849,
      "step": 22063
    },
    {
      "epoch": 5.252573945128846,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 22064
    },
    {
      "epoch": 5.252811997857526,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 22065
    },
    {
      "epoch": 5.2530500505862046,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 22066
    },
    {
      "epoch": 5.253288103314884,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 22067
    },
    {
      "epoch": 5.253526156043564,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 22068
    },
    {
      "epoch": 5.253764208772243,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 22069
    },
    {
      "epoch": 5.254002261500922,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 22070
    },
    {
      "epoch": 5.254240314229602,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 22071
    },
    {
      "epoch": 5.254478366958281,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 22072
    },
    {
      "epoch": 5.254716419686961,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 22073
    },
    {
      "epoch": 5.2549544724156405,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 22074
    },
    {
      "epoch": 5.255192525144319,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 22075
    },
    {
      "epoch": 5.255430577872999,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 22076
    },
    {
      "epoch": 5.255668630601678,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 22077
    },
    {
      "epoch": 5.255906683330358,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 22078
    },
    {
      "epoch": 5.256144736059037,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 22079
    },
    {
      "epoch": 5.256382788787716,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3329,
      "step": 22080
    },
    {
      "epoch": 5.256620841516396,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 22081
    },
    {
      "epoch": 5.256858894245076,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 22082
    },
    {
      "epoch": 5.257096946973754,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 22083
    },
    {
      "epoch": 5.257334999702434,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 22084
    },
    {
      "epoch": 5.2575730524311135,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 22085
    },
    {
      "epoch": 5.257811105159793,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 22086
    },
    {
      "epoch": 5.258049157888472,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 22087
    },
    {
      "epoch": 5.258287210617151,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3035,
      "step": 22088
    },
    {
      "epoch": 5.258525263345831,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1886,
      "step": 22089
    },
    {
      "epoch": 5.258763316074511,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 22090
    },
    {
      "epoch": 5.25900136880319,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 22091
    },
    {
      "epoch": 5.259239421531869,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 22092
    },
    {
      "epoch": 5.259477474260549,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 22093
    },
    {
      "epoch": 5.259715526989228,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3912,
      "step": 22094
    },
    {
      "epoch": 5.259953579717908,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3352,
      "step": 22095
    },
    {
      "epoch": 5.2601916324465865,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 22096
    },
    {
      "epoch": 5.260429685175266,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 22097
    },
    {
      "epoch": 5.260667737903946,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2328,
      "step": 22098
    },
    {
      "epoch": 5.260905790632625,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 22099
    },
    {
      "epoch": 5.261143843361305,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 22100
    },
    {
      "epoch": 5.261143843361305,
      "eval_runtime": 0.5286,
      "eval_samples_per_second": 1.892,
      "eval_steps_per_second": 1.892,
      "step": 22100
    },
    {
      "epoch": 5.261381896089984,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 22101
    },
    {
      "epoch": 5.261619948818663,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 22102
    },
    {
      "epoch": 5.261858001547343,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 22103
    },
    {
      "epoch": 5.2620960542760224,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3967,
      "step": 22104
    },
    {
      "epoch": 5.262334107004701,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2762,
      "step": 22105
    },
    {
      "epoch": 5.262572159733381,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 22106
    },
    {
      "epoch": 5.26281021246206,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 22107
    },
    {
      "epoch": 5.26304826519074,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 22108
    },
    {
      "epoch": 5.26328631791942,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 22109
    },
    {
      "epoch": 5.263524370648098,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 22110
    },
    {
      "epoch": 5.263762423376778,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3273,
      "step": 22111
    },
    {
      "epoch": 5.2640004761054575,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 22112
    },
    {
      "epoch": 5.264238528834137,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 22113
    },
    {
      "epoch": 5.264476581562816,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 22114
    },
    {
      "epoch": 5.264714634291495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 22115
    },
    {
      "epoch": 5.264952687020175,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1609,
      "step": 22116
    },
    {
      "epoch": 5.265190739748855,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7381,
      "step": 22117
    },
    {
      "epoch": 5.265428792477533,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 22118
    },
    {
      "epoch": 5.265666845206213,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4074,
      "step": 22119
    },
    {
      "epoch": 5.265904897934893,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 22120
    },
    {
      "epoch": 5.266142950663572,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 22121
    },
    {
      "epoch": 5.266381003392252,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 22122
    },
    {
      "epoch": 5.2666190561209305,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3901,
      "step": 22123
    },
    {
      "epoch": 5.26685710884961,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 22124
    },
    {
      "epoch": 5.26709516157829,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 22125
    },
    {
      "epoch": 5.267333214306969,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 22126
    },
    {
      "epoch": 5.267571267035648,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 22127
    },
    {
      "epoch": 5.267809319764328,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 22128
    },
    {
      "epoch": 5.268047372493007,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4445,
      "step": 22129
    },
    {
      "epoch": 5.268285425221687,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 22130
    },
    {
      "epoch": 5.268523477950366,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 22131
    },
    {
      "epoch": 5.268761530679045,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 22132
    },
    {
      "epoch": 5.268999583407725,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 22133
    },
    {
      "epoch": 5.269237636136404,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 22134
    },
    {
      "epoch": 5.269475688865084,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 22135
    },
    {
      "epoch": 5.269713741593763,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 22136
    },
    {
      "epoch": 5.269951794322442,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 22137
    },
    {
      "epoch": 5.270189847051122,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 22138
    },
    {
      "epoch": 5.2704278997798015,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3257,
      "step": 22139
    },
    {
      "epoch": 5.27066595250848,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 22140
    },
    {
      "epoch": 5.27090400523716,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 22141
    },
    {
      "epoch": 5.2711420579658395,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4177,
      "step": 22142
    },
    {
      "epoch": 5.271380110694519,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 22143
    },
    {
      "epoch": 5.271618163423199,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 22144
    },
    {
      "epoch": 5.271856216151877,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 22145
    },
    {
      "epoch": 5.272094268880557,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 22146
    },
    {
      "epoch": 5.272332321609237,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 22147
    },
    {
      "epoch": 5.272570374337916,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 22148
    },
    {
      "epoch": 5.272808427066595,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 22149
    },
    {
      "epoch": 5.2730464797952745,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 22150
    },
    {
      "epoch": 5.273284532523954,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 22151
    },
    {
      "epoch": 5.273522585252634,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 22152
    },
    {
      "epoch": 5.2737606379813124,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 22153
    },
    {
      "epoch": 5.273998690709992,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 22154
    },
    {
      "epoch": 5.274236743438672,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 22155
    },
    {
      "epoch": 5.274474796167351,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 22156
    },
    {
      "epoch": 5.274712848896031,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 22157
    },
    {
      "epoch": 5.27495090162471,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 22158
    },
    {
      "epoch": 5.275188954353389,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 22159
    },
    {
      "epoch": 5.275427007082069,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 22160
    },
    {
      "epoch": 5.275665059810748,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 22161
    },
    {
      "epoch": 5.275903112539427,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4123,
      "step": 22162
    },
    {
      "epoch": 5.276141165268107,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4119,
      "step": 22163
    },
    {
      "epoch": 5.276379217996786,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4015,
      "step": 22164
    },
    {
      "epoch": 5.276617270725466,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 22165
    },
    {
      "epoch": 5.276855323454145,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 22166
    },
    {
      "epoch": 5.277093376182824,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2047,
      "step": 22167
    },
    {
      "epoch": 5.277331428911504,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 22168
    },
    {
      "epoch": 5.2775694816401835,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 22169
    },
    {
      "epoch": 5.277807534368863,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 22170
    },
    {
      "epoch": 5.278045587097542,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 22171
    },
    {
      "epoch": 5.278283639826221,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 22172
    },
    {
      "epoch": 5.278521692554901,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 22173
    },
    {
      "epoch": 5.278759745283581,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 22174
    },
    {
      "epoch": 5.278997798012259,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 22175
    },
    {
      "epoch": 5.279235850740939,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 22176
    },
    {
      "epoch": 5.2794739034696185,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 22177
    },
    {
      "epoch": 5.279711956198298,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 22178
    },
    {
      "epoch": 5.279950008926978,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2757,
      "step": 22179
    },
    {
      "epoch": 5.2801880616556565,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2507,
      "step": 22180
    },
    {
      "epoch": 5.280426114384336,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 22181
    },
    {
      "epoch": 5.280664167113016,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 22182
    },
    {
      "epoch": 5.280902219841695,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 22183
    },
    {
      "epoch": 5.281140272570374,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 22184
    },
    {
      "epoch": 5.281378325299054,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 22185
    },
    {
      "epoch": 5.281616378027733,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 22186
    },
    {
      "epoch": 5.281854430756413,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 22187
    },
    {
      "epoch": 5.282092483485092,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 22188
    },
    {
      "epoch": 5.282330536213771,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 22189
    },
    {
      "epoch": 5.282568588942451,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 22190
    },
    {
      "epoch": 5.28280664167113,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 22191
    },
    {
      "epoch": 5.28304469439981,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 22192
    },
    {
      "epoch": 5.283282747128489,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 22193
    },
    {
      "epoch": 5.283520799857168,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 22194
    },
    {
      "epoch": 5.283758852585848,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 22195
    },
    {
      "epoch": 5.2839969053145275,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 22196
    },
    {
      "epoch": 5.284234958043206,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 22197
    },
    {
      "epoch": 5.284473010771886,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 22198
    },
    {
      "epoch": 5.284711063500565,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 22199
    },
    {
      "epoch": 5.284949116229245,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 22200
    },
    {
      "epoch": 5.284949116229245,
      "eval_runtime": 0.5235,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 22200
    },
    {
      "epoch": 5.285187168957924,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 22201
    },
    {
      "epoch": 5.285425221686603,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 22202
    },
    {
      "epoch": 5.285663274415283,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 22203
    },
    {
      "epoch": 5.2859013271439625,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 22204
    },
    {
      "epoch": 5.286139379872642,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 22205
    },
    {
      "epoch": 5.286377432601321,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 22206
    },
    {
      "epoch": 5.2866154853300005,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 22207
    },
    {
      "epoch": 5.28685353805868,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 22208
    },
    {
      "epoch": 5.28709159078736,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 22209
    },
    {
      "epoch": 5.287329643516038,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 22210
    },
    {
      "epoch": 5.287567696244718,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 22211
    },
    {
      "epoch": 5.287805748973398,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 22212
    },
    {
      "epoch": 5.288043801702077,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 22213
    },
    {
      "epoch": 5.288281854430757,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 22214
    },
    {
      "epoch": 5.2885199071594355,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 22215
    },
    {
      "epoch": 5.288757959888115,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.7615,
      "step": 22216
    },
    {
      "epoch": 5.288996012616795,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 22217
    },
    {
      "epoch": 5.289234065345474,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 22218
    },
    {
      "epoch": 5.289472118074153,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 22219
    },
    {
      "epoch": 5.289710170802833,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3604,
      "step": 22220
    },
    {
      "epoch": 5.289948223531512,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 22221
    },
    {
      "epoch": 5.290186276260192,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 22222
    },
    {
      "epoch": 5.2904243289888715,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 22223
    },
    {
      "epoch": 5.29066238171755,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 22224
    },
    {
      "epoch": 5.29090043444623,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 22225
    },
    {
      "epoch": 5.291138487174909,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4815,
      "step": 22226
    },
    {
      "epoch": 5.291376539903589,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 22227
    },
    {
      "epoch": 5.291614592632268,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 22228
    },
    {
      "epoch": 5.291852645360947,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 22229
    },
    {
      "epoch": 5.292090698089627,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 22230
    },
    {
      "epoch": 5.292328750818307,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 22231
    },
    {
      "epoch": 5.292566803546985,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 22232
    },
    {
      "epoch": 5.292804856275665,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 22233
    },
    {
      "epoch": 5.2930429090043445,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 22234
    },
    {
      "epoch": 5.293280961733024,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8902,
      "step": 22235
    },
    {
      "epoch": 5.293519014461703,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 22236
    },
    {
      "epoch": 5.293757067190382,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3913,
      "step": 22237
    },
    {
      "epoch": 5.293995119919062,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 22238
    },
    {
      "epoch": 5.294233172647742,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 22239
    },
    {
      "epoch": 5.294471225376421,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 22240
    },
    {
      "epoch": 5.2947092781051,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 22241
    },
    {
      "epoch": 5.2949473308337796,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 22242
    },
    {
      "epoch": 5.295185383562459,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3039,
      "step": 22243
    },
    {
      "epoch": 5.295423436291139,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8828,
      "step": 22244
    },
    {
      "epoch": 5.2956614890198175,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 22245
    },
    {
      "epoch": 5.295899541748497,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 22246
    },
    {
      "epoch": 5.296137594477177,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.5207,
      "step": 22247
    },
    {
      "epoch": 5.296375647205856,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 22248
    },
    {
      "epoch": 5.296613699934536,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 22249
    },
    {
      "epoch": 5.296851752663215,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 22250
    },
    {
      "epoch": 5.297089805391894,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 22251
    },
    {
      "epoch": 5.297327858120574,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 22252
    },
    {
      "epoch": 5.297565910849253,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 22253
    },
    {
      "epoch": 5.297803963577932,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 22254
    },
    {
      "epoch": 5.298042016306612,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 22255
    },
    {
      "epoch": 5.298280069035291,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 22256
    },
    {
      "epoch": 5.298518121763971,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 22257
    },
    {
      "epoch": 5.298756174492651,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 22258
    },
    {
      "epoch": 5.298994227221329,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 22259
    },
    {
      "epoch": 5.299232279950009,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 22260
    },
    {
      "epoch": 5.2994703326786885,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 22261
    },
    {
      "epoch": 5.299708385407368,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 22262
    },
    {
      "epoch": 5.299946438136047,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 22263
    },
    {
      "epoch": 5.300184490864726,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 22264
    },
    {
      "epoch": 5.300422543593406,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 22265
    },
    {
      "epoch": 5.300660596322086,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2627,
      "step": 22266
    },
    {
      "epoch": 5.300898649050764,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 22267
    },
    {
      "epoch": 5.301136701779444,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 22268
    },
    {
      "epoch": 5.301374754508124,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 22269
    },
    {
      "epoch": 5.301612807236803,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 22270
    },
    {
      "epoch": 5.301850859965482,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2393,
      "step": 22271
    },
    {
      "epoch": 5.3020889126941615,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 22272
    },
    {
      "epoch": 5.302326965422841,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 22273
    },
    {
      "epoch": 5.302565018151521,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 22274
    },
    {
      "epoch": 5.3028030708802,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 22275
    },
    {
      "epoch": 5.303041123608879,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 22276
    },
    {
      "epoch": 5.303279176337559,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 22277
    },
    {
      "epoch": 5.303517229066238,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 22278
    },
    {
      "epoch": 5.303755281794918,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 22279
    },
    {
      "epoch": 5.303993334523597,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 22280
    },
    {
      "epoch": 5.304231387252276,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 22281
    },
    {
      "epoch": 5.304469439980956,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 22282
    },
    {
      "epoch": 5.304707492709635,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 22283
    },
    {
      "epoch": 5.304945545438315,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 22284
    },
    {
      "epoch": 5.305183598166994,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 22285
    },
    {
      "epoch": 5.305421650895673,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 22286
    },
    {
      "epoch": 5.305659703624353,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 22287
    },
    {
      "epoch": 5.3058977563530325,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 22288
    },
    {
      "epoch": 5.306135809081711,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 22289
    },
    {
      "epoch": 5.306373861810391,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 22290
    },
    {
      "epoch": 5.30661191453907,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 22291
    },
    {
      "epoch": 5.30684996726775,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 22292
    },
    {
      "epoch": 5.30708801999643,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 22293
    },
    {
      "epoch": 5.307326072725108,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 22294
    },
    {
      "epoch": 5.307564125453788,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 22295
    },
    {
      "epoch": 5.307802178182468,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 22296
    },
    {
      "epoch": 5.308040230911147,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 22297
    },
    {
      "epoch": 5.308278283639826,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 22298
    },
    {
      "epoch": 5.3085163363685055,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 22299
    },
    {
      "epoch": 5.308754389097185,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1939,
      "step": 22300
    },
    {
      "epoch": 5.308754389097185,
      "eval_runtime": 0.5427,
      "eval_samples_per_second": 1.843,
      "eval_steps_per_second": 1.843,
      "step": 22300
    },
    {
      "epoch": 5.308992441825865,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 22301
    },
    {
      "epoch": 5.309230494554543,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 22302
    },
    {
      "epoch": 5.309468547283223,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 22303
    },
    {
      "epoch": 5.309706600011903,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 22304
    },
    {
      "epoch": 5.309944652740582,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7202,
      "step": 22305
    },
    {
      "epoch": 5.310182705469262,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 22306
    },
    {
      "epoch": 5.310420758197941,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 22307
    },
    {
      "epoch": 5.31065881092662,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 22308
    },
    {
      "epoch": 5.3108968636553,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 22309
    },
    {
      "epoch": 5.311134916383979,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 22310
    },
    {
      "epoch": 5.311372969112658,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.791,
      "step": 22311
    },
    {
      "epoch": 5.311611021841338,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 22312
    },
    {
      "epoch": 5.311849074570017,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 22313
    },
    {
      "epoch": 5.312087127298697,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 22314
    },
    {
      "epoch": 5.312325180027376,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 22315
    },
    {
      "epoch": 5.312563232756055,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 22316
    },
    {
      "epoch": 5.312801285484735,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 22317
    },
    {
      "epoch": 5.3130393382134145,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 22318
    },
    {
      "epoch": 5.313277390942094,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 22319
    },
    {
      "epoch": 5.313515443670773,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 22320
    },
    {
      "epoch": 5.313753496399452,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 22321
    },
    {
      "epoch": 5.313991549128132,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 22322
    },
    {
      "epoch": 5.314229601856812,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 22323
    },
    {
      "epoch": 5.31446765458549,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 22324
    },
    {
      "epoch": 5.31470570731417,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 22325
    },
    {
      "epoch": 5.3149437600428495,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 22326
    },
    {
      "epoch": 5.315181812771529,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 22327
    },
    {
      "epoch": 5.315419865500209,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2832,
      "step": 22328
    },
    {
      "epoch": 5.3156579182288874,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 22329
    },
    {
      "epoch": 5.315895970957567,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3163,
      "step": 22330
    },
    {
      "epoch": 5.316134023686247,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 22331
    },
    {
      "epoch": 5.316372076414926,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 22332
    },
    {
      "epoch": 5.316610129143605,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 22333
    },
    {
      "epoch": 5.316848181872285,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 22334
    },
    {
      "epoch": 5.317086234600964,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 22335
    },
    {
      "epoch": 5.317324287329644,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 22336
    },
    {
      "epoch": 5.3175623400583225,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 22337
    },
    {
      "epoch": 5.317800392787002,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 22338
    },
    {
      "epoch": 5.318038445515682,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 22339
    },
    {
      "epoch": 5.318276498244361,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 22340
    },
    {
      "epoch": 5.318514550973041,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 22341
    },
    {
      "epoch": 5.31875260370172,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 22342
    },
    {
      "epoch": 5.318990656430399,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 22343
    },
    {
      "epoch": 5.319228709159079,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 22344
    },
    {
      "epoch": 5.3194667618877585,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 22345
    },
    {
      "epoch": 5.319704814616437,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 22346
    },
    {
      "epoch": 5.319942867345117,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 22347
    },
    {
      "epoch": 5.320180920073796,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 22348
    },
    {
      "epoch": 5.320418972802476,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2846,
      "step": 22349
    },
    {
      "epoch": 5.320657025531155,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 22350
    },
    {
      "epoch": 5.320895078259834,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 22351
    },
    {
      "epoch": 5.321133130988514,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 22352
    },
    {
      "epoch": 5.3213711837171935,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 22353
    },
    {
      "epoch": 5.321609236445873,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 22354
    },
    {
      "epoch": 5.321847289174552,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 22355
    },
    {
      "epoch": 5.3220853419032315,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 22356
    },
    {
      "epoch": 5.322323394631911,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2134,
      "step": 22357
    },
    {
      "epoch": 5.322561447360591,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 22358
    },
    {
      "epoch": 5.322799500089269,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4897,
      "step": 22359
    },
    {
      "epoch": 5.323037552817949,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 22360
    },
    {
      "epoch": 5.323275605546629,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 22361
    },
    {
      "epoch": 5.323513658275308,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 22362
    },
    {
      "epoch": 5.323751711003988,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 22363
    },
    {
      "epoch": 5.3239897637326665,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 22364
    },
    {
      "epoch": 5.324227816461346,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 22365
    },
    {
      "epoch": 5.324465869190026,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 22366
    },
    {
      "epoch": 5.324703921918705,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 22367
    },
    {
      "epoch": 5.324941974647384,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 22368
    },
    {
      "epoch": 5.325180027376064,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 22369
    },
    {
      "epoch": 5.325418080104743,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 22370
    },
    {
      "epoch": 5.325656132833423,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 22371
    },
    {
      "epoch": 5.325894185562102,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.3189,
      "step": 22372
    },
    {
      "epoch": 5.326132238290781,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 22373
    },
    {
      "epoch": 5.326370291019461,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 22374
    },
    {
      "epoch": 5.32660834374814,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 22375
    },
    {
      "epoch": 5.32684639647682,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 22376
    },
    {
      "epoch": 5.327084449205499,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 22377
    },
    {
      "epoch": 5.327322501934178,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 22378
    },
    {
      "epoch": 5.327560554662858,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3374,
      "step": 22379
    },
    {
      "epoch": 5.3277986073915375,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 22380
    },
    {
      "epoch": 5.328036660120216,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 22381
    },
    {
      "epoch": 5.328274712848896,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 22382
    },
    {
      "epoch": 5.3285127655775755,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 22383
    },
    {
      "epoch": 5.328750818306255,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 22384
    },
    {
      "epoch": 5.328988871034934,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 22385
    },
    {
      "epoch": 5.329226923763613,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8013,
      "step": 22386
    },
    {
      "epoch": 5.329464976492293,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 22387
    },
    {
      "epoch": 5.329703029220973,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 22388
    },
    {
      "epoch": 5.329941081949652,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 22389
    },
    {
      "epoch": 5.330179134678331,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3607,
      "step": 22390
    },
    {
      "epoch": 5.3304171874070105,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 22391
    },
    {
      "epoch": 5.33065524013569,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 22392
    },
    {
      "epoch": 5.33089329286437,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 22393
    },
    {
      "epoch": 5.3311313455930485,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 22394
    },
    {
      "epoch": 5.331369398321728,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 22395
    },
    {
      "epoch": 5.331607451050408,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8048,
      "step": 22396
    },
    {
      "epoch": 5.331845503779087,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 22397
    },
    {
      "epoch": 5.332083556507767,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.6845,
      "step": 22398
    },
    {
      "epoch": 5.332321609236446,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 22399
    },
    {
      "epoch": 5.332559661965125,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 22400
    },
    {
      "epoch": 5.332559661965125,
      "eval_runtime": 0.5283,
      "eval_samples_per_second": 1.893,
      "eval_steps_per_second": 1.893,
      "step": 22400
    },
    {
      "epoch": 5.332797714693805,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 22401
    },
    {
      "epoch": 5.333035767422484,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 22402
    },
    {
      "epoch": 5.333273820151163,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 22403
    },
    {
      "epoch": 5.333511872879843,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7372,
      "step": 22404
    },
    {
      "epoch": 5.333749925608522,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 22405
    },
    {
      "epoch": 5.333987978337202,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 22406
    },
    {
      "epoch": 5.334226031065882,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 22407
    },
    {
      "epoch": 5.33446408379456,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 22408
    },
    {
      "epoch": 5.33470213652324,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 22409
    },
    {
      "epoch": 5.3349401892519195,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8524,
      "step": 22410
    },
    {
      "epoch": 5.335178241980599,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 22411
    },
    {
      "epoch": 5.335416294709278,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2784,
      "step": 22412
    },
    {
      "epoch": 5.335654347437957,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 22413
    },
    {
      "epoch": 5.335892400166637,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 22414
    },
    {
      "epoch": 5.336130452895317,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 22415
    },
    {
      "epoch": 5.336368505623995,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 22416
    },
    {
      "epoch": 5.336606558352675,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 22417
    },
    {
      "epoch": 5.3368446110813546,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 22418
    },
    {
      "epoch": 5.337082663810034,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 22419
    },
    {
      "epoch": 5.337320716538713,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 22420
    },
    {
      "epoch": 5.3375587692673925,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 22421
    },
    {
      "epoch": 5.337796821996072,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 22422
    },
    {
      "epoch": 5.338034874724752,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 22423
    },
    {
      "epoch": 5.338272927453431,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 22424
    },
    {
      "epoch": 5.33851098018211,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 22425
    },
    {
      "epoch": 5.33874903291079,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 22426
    },
    {
      "epoch": 5.338987085639469,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 22427
    },
    {
      "epoch": 5.339225138368149,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3327,
      "step": 22428
    },
    {
      "epoch": 5.3394631910968275,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 22429
    },
    {
      "epoch": 5.339701243825507,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 22430
    },
    {
      "epoch": 5.339939296554187,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 22431
    },
    {
      "epoch": 5.340177349282866,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 22432
    },
    {
      "epoch": 5.340415402011546,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 22433
    },
    {
      "epoch": 5.340653454740225,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 22434
    },
    {
      "epoch": 5.340891507468904,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 22435
    },
    {
      "epoch": 5.341129560197584,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 22436
    },
    {
      "epoch": 5.3413676129262635,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 22437
    },
    {
      "epoch": 5.341605665654942,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 22438
    },
    {
      "epoch": 5.341843718383622,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.5815,
      "step": 22439
    },
    {
      "epoch": 5.342081771112301,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 22440
    },
    {
      "epoch": 5.342319823840981,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.4046,
      "step": 22441
    },
    {
      "epoch": 5.342557876569661,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 22442
    },
    {
      "epoch": 5.342795929298339,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1485,
      "step": 22443
    },
    {
      "epoch": 5.343033982027019,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3503,
      "step": 22444
    },
    {
      "epoch": 5.343272034755699,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 22445
    },
    {
      "epoch": 5.343510087484378,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 22446
    },
    {
      "epoch": 5.343748140213057,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 22447
    },
    {
      "epoch": 5.3439861929417365,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.614,
      "step": 22448
    },
    {
      "epoch": 5.344224245670416,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 22449
    },
    {
      "epoch": 5.344462298399096,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 22450
    },
    {
      "epoch": 5.344700351127774,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 22451
    },
    {
      "epoch": 5.344938403856454,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 22452
    },
    {
      "epoch": 5.345176456585134,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 22453
    },
    {
      "epoch": 5.345414509313813,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 22454
    },
    {
      "epoch": 5.345652562042492,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3019,
      "step": 22455
    },
    {
      "epoch": 5.345890614771172,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 22456
    },
    {
      "epoch": 5.346128667499851,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 22457
    },
    {
      "epoch": 5.346366720228531,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 22458
    },
    {
      "epoch": 5.34660477295721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 22459
    },
    {
      "epoch": 5.346842825685889,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 22460
    },
    {
      "epoch": 5.347080878414569,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 22461
    },
    {
      "epoch": 5.347318931143248,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 22462
    },
    {
      "epoch": 5.347556983871928,
      "grad_norm": 2.71875,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 22463
    },
    {
      "epoch": 5.347795036600607,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 22464
    },
    {
      "epoch": 5.348033089329286,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 22465
    },
    {
      "epoch": 5.348271142057966,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 22466
    },
    {
      "epoch": 5.348509194786645,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 22467
    },
    {
      "epoch": 5.348747247515325,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 22468
    },
    {
      "epoch": 5.348985300244004,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 22469
    },
    {
      "epoch": 5.349223352972683,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 22470
    },
    {
      "epoch": 5.349461405701363,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 22471
    },
    {
      "epoch": 5.349699458430043,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 22472
    },
    {
      "epoch": 5.349937511158721,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 22473
    },
    {
      "epoch": 5.350175563887401,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 22474
    },
    {
      "epoch": 5.3504136166160805,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.4955,
      "step": 22475
    },
    {
      "epoch": 5.35065166934476,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 22476
    },
    {
      "epoch": 5.35088972207344,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 22477
    },
    {
      "epoch": 5.351127774802118,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 22478
    },
    {
      "epoch": 5.351365827530798,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 22479
    },
    {
      "epoch": 5.351603880259478,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 22480
    },
    {
      "epoch": 5.351841932988157,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 22481
    },
    {
      "epoch": 5.352079985716836,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1923,
      "step": 22482
    },
    {
      "epoch": 5.352318038445516,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4718,
      "step": 22483
    },
    {
      "epoch": 5.352556091174195,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 22484
    },
    {
      "epoch": 5.352794143902875,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 22485
    },
    {
      "epoch": 5.3530321966315535,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 22486
    },
    {
      "epoch": 5.353270249360233,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 22487
    },
    {
      "epoch": 5.353508302088913,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 22488
    },
    {
      "epoch": 5.353746354817592,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4221,
      "step": 22489
    },
    {
      "epoch": 5.353984407546271,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 22490
    },
    {
      "epoch": 5.354222460274951,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4655,
      "step": 22491
    },
    {
      "epoch": 5.35446051300363,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 22492
    },
    {
      "epoch": 5.35469856573231,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 22493
    },
    {
      "epoch": 5.3549366184609895,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 22494
    },
    {
      "epoch": 5.355174671189668,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 22495
    },
    {
      "epoch": 5.355412723918348,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 22496
    },
    {
      "epoch": 5.355650776647027,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 22497
    },
    {
      "epoch": 5.355888829375707,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 22498
    },
    {
      "epoch": 5.356126882104386,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 22499
    },
    {
      "epoch": 5.356364934833065,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2946,
      "step": 22500
    },
    {
      "epoch": 5.356364934833065,
      "eval_runtime": 0.5116,
      "eval_samples_per_second": 1.955,
      "eval_steps_per_second": 1.955,
      "step": 22500
    },
    {
      "epoch": 5.356602987561745,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 22501
    },
    {
      "epoch": 5.3568410402904245,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5036,
      "step": 22502
    },
    {
      "epoch": 5.357079093019104,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 22503
    },
    {
      "epoch": 5.357317145747783,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 22504
    },
    {
      "epoch": 5.357555198476462,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 22505
    },
    {
      "epoch": 5.357793251205142,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 22506
    },
    {
      "epoch": 5.358031303933822,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 22507
    },
    {
      "epoch": 5.3582693566625,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 22508
    },
    {
      "epoch": 5.35850740939118,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 22509
    },
    {
      "epoch": 5.35874546211986,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 22510
    },
    {
      "epoch": 5.358983514848539,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 22511
    },
    {
      "epoch": 5.359221567577219,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.4438,
      "step": 22512
    },
    {
      "epoch": 5.3594596203058975,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 22513
    },
    {
      "epoch": 5.359697673034577,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 22514
    },
    {
      "epoch": 5.359935725763257,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 22515
    },
    {
      "epoch": 5.360173778491936,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 22516
    },
    {
      "epoch": 5.360411831220615,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 22517
    },
    {
      "epoch": 5.360649883949295,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 22518
    },
    {
      "epoch": 5.360887936677974,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 22519
    },
    {
      "epoch": 5.361125989406654,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 22520
    },
    {
      "epoch": 5.361364042135333,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 22521
    },
    {
      "epoch": 5.361602094864012,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 22522
    },
    {
      "epoch": 5.361840147592692,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 22523
    },
    {
      "epoch": 5.362078200321371,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2879,
      "step": 22524
    },
    {
      "epoch": 5.362316253050051,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3759,
      "step": 22525
    },
    {
      "epoch": 5.36255430577873,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 22526
    },
    {
      "epoch": 5.362792358507409,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 22527
    },
    {
      "epoch": 5.363030411236089,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2511,
      "step": 22528
    },
    {
      "epoch": 5.3632684639647685,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 22529
    },
    {
      "epoch": 5.363506516693447,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 22530
    },
    {
      "epoch": 5.363744569422127,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 22531
    },
    {
      "epoch": 5.3639826221508065,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0239,
      "step": 22532
    },
    {
      "epoch": 5.364220674879486,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 22533
    },
    {
      "epoch": 5.364458727608165,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 22534
    },
    {
      "epoch": 5.364696780336844,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 22535
    },
    {
      "epoch": 5.364934833065524,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 22536
    },
    {
      "epoch": 5.365172885794204,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 22537
    },
    {
      "epoch": 5.365410938522883,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 22538
    },
    {
      "epoch": 5.365648991251562,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 22539
    },
    {
      "epoch": 5.3658870439802415,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 22540
    },
    {
      "epoch": 5.366125096708921,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 22541
    },
    {
      "epoch": 5.366363149437601,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 22542
    },
    {
      "epoch": 5.3666012021662794,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 22543
    },
    {
      "epoch": 5.366839254894959,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 22544
    },
    {
      "epoch": 5.367077307623639,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 22545
    },
    {
      "epoch": 5.367315360352318,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 22546
    },
    {
      "epoch": 5.367553413080998,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 22547
    },
    {
      "epoch": 5.367791465809677,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 22548
    },
    {
      "epoch": 5.368029518538356,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 22549
    },
    {
      "epoch": 5.368267571267036,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 22550
    },
    {
      "epoch": 5.368505623995715,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0561,
      "step": 22551
    },
    {
      "epoch": 5.368743676724394,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 22552
    },
    {
      "epoch": 5.368981729453074,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 22553
    },
    {
      "epoch": 5.369219782181753,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 22554
    },
    {
      "epoch": 5.369457834910433,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 22555
    },
    {
      "epoch": 5.369695887639112,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2697,
      "step": 22556
    },
    {
      "epoch": 5.369933940367791,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 22557
    },
    {
      "epoch": 5.370171993096471,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 22558
    },
    {
      "epoch": 5.3704100458251505,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 22559
    },
    {
      "epoch": 5.37064809855383,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7183,
      "step": 22560
    },
    {
      "epoch": 5.370886151282509,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 22561
    },
    {
      "epoch": 5.371124204011188,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 22562
    },
    {
      "epoch": 5.371362256739868,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 22563
    },
    {
      "epoch": 5.371600309468548,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 22564
    },
    {
      "epoch": 5.371838362197226,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 22565
    },
    {
      "epoch": 5.372076414925906,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 22566
    },
    {
      "epoch": 5.3723144676545855,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 22567
    },
    {
      "epoch": 5.372552520383265,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 22568
    },
    {
      "epoch": 5.372790573111944,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 22569
    },
    {
      "epoch": 5.3730286258406235,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 22570
    },
    {
      "epoch": 5.373266678569303,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 22571
    },
    {
      "epoch": 5.373504731297983,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 22572
    },
    {
      "epoch": 5.373742784026662,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 22573
    },
    {
      "epoch": 5.373980836755341,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 22574
    },
    {
      "epoch": 5.374218889484021,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 22575
    },
    {
      "epoch": 5.3744569422127,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 22576
    },
    {
      "epoch": 5.37469499494138,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 22577
    },
    {
      "epoch": 5.3749330476700585,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 22578
    },
    {
      "epoch": 5.375171100398738,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 22579
    },
    {
      "epoch": 5.375409153127418,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 22580
    },
    {
      "epoch": 5.375647205856097,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 22581
    },
    {
      "epoch": 5.375885258584777,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 22582
    },
    {
      "epoch": 5.376123311313456,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 22583
    },
    {
      "epoch": 5.376361364042135,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4397,
      "step": 22584
    },
    {
      "epoch": 5.376599416770815,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 22585
    },
    {
      "epoch": 5.3768374694994945,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 22586
    },
    {
      "epoch": 5.377075522228173,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 22587
    },
    {
      "epoch": 5.377313574956853,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8766,
      "step": 22588
    },
    {
      "epoch": 5.377551627685532,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 22589
    },
    {
      "epoch": 5.377789680414212,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 22590
    },
    {
      "epoch": 5.378027733142892,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 22591
    },
    {
      "epoch": 5.37826578587157,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 22592
    },
    {
      "epoch": 5.37850383860025,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3238,
      "step": 22593
    },
    {
      "epoch": 5.3787418913289295,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 22594
    },
    {
      "epoch": 5.378979944057609,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 22595
    },
    {
      "epoch": 5.379217996786288,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 22596
    },
    {
      "epoch": 5.3794560495149675,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 22597
    },
    {
      "epoch": 5.379694102243647,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 22598
    },
    {
      "epoch": 5.379932154972327,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 22599
    },
    {
      "epoch": 5.380170207701005,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 22600
    },
    {
      "epoch": 5.380170207701005,
      "eval_runtime": 0.5193,
      "eval_samples_per_second": 1.926,
      "eval_steps_per_second": 1.926,
      "step": 22600
    },
    {
      "epoch": 5.380408260429685,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 22601
    },
    {
      "epoch": 5.380646313158365,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 22602
    },
    {
      "epoch": 5.380884365887044,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 22603
    },
    {
      "epoch": 5.381122418615723,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 22604
    },
    {
      "epoch": 5.3813604713444025,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 22605
    },
    {
      "epoch": 5.381598524073082,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 22606
    },
    {
      "epoch": 5.381836576801762,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 22607
    },
    {
      "epoch": 5.382074629530441,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 22608
    },
    {
      "epoch": 5.38231268225912,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4471,
      "step": 22609
    },
    {
      "epoch": 5.3825507349878,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 22610
    },
    {
      "epoch": 5.382788787716479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7069,
      "step": 22611
    },
    {
      "epoch": 5.383026840445159,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 22612
    },
    {
      "epoch": 5.383264893173838,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 22613
    },
    {
      "epoch": 5.383502945902517,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4547,
      "step": 22614
    },
    {
      "epoch": 5.383740998631197,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 22615
    },
    {
      "epoch": 5.383979051359876,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 22616
    },
    {
      "epoch": 5.384217104088556,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 22617
    },
    {
      "epoch": 5.384455156817235,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.6033,
      "step": 22618
    },
    {
      "epoch": 5.384693209545914,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 22619
    },
    {
      "epoch": 5.384931262274594,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 22620
    },
    {
      "epoch": 5.385169315003274,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 22621
    },
    {
      "epoch": 5.385407367731952,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2695,
      "step": 22622
    },
    {
      "epoch": 5.385645420460632,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 22623
    },
    {
      "epoch": 5.3858834731893115,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 22624
    },
    {
      "epoch": 5.386121525917991,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 22625
    },
    {
      "epoch": 5.386359578646671,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 22626
    },
    {
      "epoch": 5.386597631375349,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 22627
    },
    {
      "epoch": 5.386835684104029,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3842,
      "step": 22628
    },
    {
      "epoch": 5.387073736832709,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 22629
    },
    {
      "epoch": 5.387311789561388,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 22630
    },
    {
      "epoch": 5.387549842290067,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 22631
    },
    {
      "epoch": 5.3877878950187466,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 22632
    },
    {
      "epoch": 5.388025947747426,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 22633
    },
    {
      "epoch": 5.388264000476106,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 22634
    },
    {
      "epoch": 5.3885020532047845,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 22635
    },
    {
      "epoch": 5.388740105933464,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 22636
    },
    {
      "epoch": 5.388978158662144,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4457,
      "step": 22637
    },
    {
      "epoch": 5.389216211390823,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 22638
    },
    {
      "epoch": 5.389454264119502,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 22639
    },
    {
      "epoch": 5.389692316848182,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 22640
    },
    {
      "epoch": 5.389930369576861,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 22641
    },
    {
      "epoch": 5.390168422305541,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 22642
    },
    {
      "epoch": 5.39040647503422,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 22643
    },
    {
      "epoch": 5.390644527762899,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 22644
    },
    {
      "epoch": 5.390882580491579,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 22645
    },
    {
      "epoch": 5.391120633220258,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.5886,
      "step": 22646
    },
    {
      "epoch": 5.391358685948938,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 22647
    },
    {
      "epoch": 5.391596738677617,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 22648
    },
    {
      "epoch": 5.391834791406296,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8369,
      "step": 22649
    },
    {
      "epoch": 5.392072844134976,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4328,
      "step": 22650
    },
    {
      "epoch": 5.3923108968636555,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 22651
    },
    {
      "epoch": 5.392548949592335,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 22652
    },
    {
      "epoch": 5.392787002321014,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 22653
    },
    {
      "epoch": 5.393025055049693,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 22654
    },
    {
      "epoch": 5.393263107778373,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 22655
    },
    {
      "epoch": 5.393501160507053,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2947,
      "step": 22656
    },
    {
      "epoch": 5.393739213235731,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 22657
    },
    {
      "epoch": 5.393977265964411,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 22658
    },
    {
      "epoch": 5.394215318693091,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 22659
    },
    {
      "epoch": 5.39445337142177,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 22660
    },
    {
      "epoch": 5.39469142415045,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 22661
    },
    {
      "epoch": 5.3949294768791285,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 22662
    },
    {
      "epoch": 5.395167529607808,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3132,
      "step": 22663
    },
    {
      "epoch": 5.395405582336488,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 22664
    },
    {
      "epoch": 5.395643635065167,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 22665
    },
    {
      "epoch": 5.395881687793846,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 22666
    },
    {
      "epoch": 5.396119740522526,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 22667
    },
    {
      "epoch": 5.396357793251205,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 22668
    },
    {
      "epoch": 5.396595845979885,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 22669
    },
    {
      "epoch": 5.396833898708564,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 22670
    },
    {
      "epoch": 5.397071951437243,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 22671
    },
    {
      "epoch": 5.397310004165923,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 22672
    },
    {
      "epoch": 5.397548056894602,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 22673
    },
    {
      "epoch": 5.397786109623281,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 22674
    },
    {
      "epoch": 5.398024162351961,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 22675
    },
    {
      "epoch": 5.39826221508064,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 22676
    },
    {
      "epoch": 5.39850026780932,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 22677
    },
    {
      "epoch": 5.3987383205379995,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 22678
    },
    {
      "epoch": 5.398976373266678,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 22679
    },
    {
      "epoch": 5.399214425995358,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 22680
    },
    {
      "epoch": 5.399452478724037,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 22681
    },
    {
      "epoch": 5.399690531452717,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 22682
    },
    {
      "epoch": 5.399928584181396,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 22683
    },
    {
      "epoch": 5.400166636910075,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 22684
    },
    {
      "epoch": 5.400404689638755,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 22685
    },
    {
      "epoch": 5.400642742367435,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 22686
    },
    {
      "epoch": 5.400880795096114,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 22687
    },
    {
      "epoch": 5.401118847824793,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 22688
    },
    {
      "epoch": 5.4013569005534725,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 22689
    },
    {
      "epoch": 5.401594953282152,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 22690
    },
    {
      "epoch": 5.401833006010832,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 22691
    },
    {
      "epoch": 5.40207105873951,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 22692
    },
    {
      "epoch": 5.40230911146819,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 22693
    },
    {
      "epoch": 5.40254716419687,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 22694
    },
    {
      "epoch": 5.402785216925549,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 22695
    },
    {
      "epoch": 5.403023269654229,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 22696
    },
    {
      "epoch": 5.403261322382908,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 22697
    },
    {
      "epoch": 5.403499375111587,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 22698
    },
    {
      "epoch": 5.403737427840267,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 22699
    },
    {
      "epoch": 5.403975480568946,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 22700
    },
    {
      "epoch": 5.403975480568946,
      "eval_runtime": 0.5206,
      "eval_samples_per_second": 1.921,
      "eval_steps_per_second": 1.921,
      "step": 22700
    },
    {
      "epoch": 5.404213533297625,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 22701
    },
    {
      "epoch": 5.404451586026305,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 22702
    },
    {
      "epoch": 5.404689638754984,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 22703
    },
    {
      "epoch": 5.404927691483664,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 22704
    },
    {
      "epoch": 5.405165744212343,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 22705
    },
    {
      "epoch": 5.405403796941022,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2674,
      "step": 22706
    },
    {
      "epoch": 5.405641849669702,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 22707
    },
    {
      "epoch": 5.4058799023983815,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 22708
    },
    {
      "epoch": 5.406117955127061,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 22709
    },
    {
      "epoch": 5.40635600785574,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 22710
    },
    {
      "epoch": 5.406594060584419,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 22711
    },
    {
      "epoch": 5.406832113313099,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 22712
    },
    {
      "epoch": 5.407070166041779,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4364,
      "step": 22713
    },
    {
      "epoch": 5.407308218770457,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.4595,
      "step": 22714
    },
    {
      "epoch": 5.407546271499137,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 22715
    },
    {
      "epoch": 5.4077843242278165,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 22716
    },
    {
      "epoch": 5.408022376956496,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3179,
      "step": 22717
    },
    {
      "epoch": 5.408260429685175,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 22718
    },
    {
      "epoch": 5.4084984824138544,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 22719
    },
    {
      "epoch": 5.408736535142534,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3687,
      "step": 22720
    },
    {
      "epoch": 5.408974587871214,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4401,
      "step": 22721
    },
    {
      "epoch": 5.409212640599893,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 22722
    },
    {
      "epoch": 5.409450693328572,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 22723
    },
    {
      "epoch": 5.409688746057252,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.0497,
      "step": 22724
    },
    {
      "epoch": 5.409926798785931,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 22725
    },
    {
      "epoch": 5.410164851514611,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 22726
    },
    {
      "epoch": 5.4104029042432895,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2663,
      "step": 22727
    },
    {
      "epoch": 5.410640956971969,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 22728
    },
    {
      "epoch": 5.410879009700649,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 22729
    },
    {
      "epoch": 5.411117062429328,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 22730
    },
    {
      "epoch": 5.411355115158008,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 22731
    },
    {
      "epoch": 5.411593167886687,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 22732
    },
    {
      "epoch": 5.411831220615366,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 22733
    },
    {
      "epoch": 5.412069273344046,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 22734
    },
    {
      "epoch": 5.4123073260727255,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6514,
      "step": 22735
    },
    {
      "epoch": 5.412545378801404,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 22736
    },
    {
      "epoch": 5.412783431530084,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 22737
    },
    {
      "epoch": 5.413021484258763,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 22738
    },
    {
      "epoch": 5.413259536987443,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 22739
    },
    {
      "epoch": 5.413497589716122,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 22740
    },
    {
      "epoch": 5.413735642444801,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 22741
    },
    {
      "epoch": 5.413973695173481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 22742
    },
    {
      "epoch": 5.4142117479021605,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 22743
    },
    {
      "epoch": 5.41444980063084,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 22744
    },
    {
      "epoch": 5.414687853359519,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 22745
    },
    {
      "epoch": 5.4149259060881985,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 22746
    },
    {
      "epoch": 5.415163958816878,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 22747
    },
    {
      "epoch": 5.415402011545558,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 22748
    },
    {
      "epoch": 5.415640064274236,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 22749
    },
    {
      "epoch": 5.415878117002916,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 22750
    },
    {
      "epoch": 5.416116169731596,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 22751
    },
    {
      "epoch": 5.416354222460275,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 22752
    },
    {
      "epoch": 5.416592275188954,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.327,
      "step": 22753
    },
    {
      "epoch": 5.4168303279176335,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 22754
    },
    {
      "epoch": 5.417068380646313,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 22755
    },
    {
      "epoch": 5.417306433374993,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 22756
    },
    {
      "epoch": 5.417544486103672,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 22757
    },
    {
      "epoch": 5.417782538832351,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 22758
    },
    {
      "epoch": 5.418020591561031,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 22759
    },
    {
      "epoch": 5.41825864428971,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 22760
    },
    {
      "epoch": 5.41849669701839,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 22761
    },
    {
      "epoch": 5.418734749747069,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1948,
      "step": 22762
    },
    {
      "epoch": 5.418972802475748,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 22763
    },
    {
      "epoch": 5.419210855204428,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 22764
    },
    {
      "epoch": 5.419448907933107,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 22765
    },
    {
      "epoch": 5.419686960661787,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 22766
    },
    {
      "epoch": 5.419925013390466,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 22767
    },
    {
      "epoch": 5.420163066119145,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 22768
    },
    {
      "epoch": 5.420401118847825,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 22769
    },
    {
      "epoch": 5.4206391715765045,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 22770
    },
    {
      "epoch": 5.420877224305183,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 22771
    },
    {
      "epoch": 5.421115277033863,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 22772
    },
    {
      "epoch": 5.4213533297625425,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 22773
    },
    {
      "epoch": 5.421591382491222,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 22774
    },
    {
      "epoch": 5.421829435219901,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 22775
    },
    {
      "epoch": 5.42206748794858,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.915,
      "step": 22776
    },
    {
      "epoch": 5.42230554067726,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 22777
    },
    {
      "epoch": 5.42254359340594,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 22778
    },
    {
      "epoch": 5.422781646134619,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 22779
    },
    {
      "epoch": 5.423019698863298,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 22780
    },
    {
      "epoch": 5.4232577515919775,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 22781
    },
    {
      "epoch": 5.423495804320657,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 22782
    },
    {
      "epoch": 5.423733857049337,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 22783
    },
    {
      "epoch": 5.4239719097780155,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 22784
    },
    {
      "epoch": 5.424209962506695,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 22785
    },
    {
      "epoch": 5.424448015235375,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 22786
    },
    {
      "epoch": 5.424686067964054,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 22787
    },
    {
      "epoch": 5.424924120692733,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 22788
    },
    {
      "epoch": 5.425162173421413,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 22789
    },
    {
      "epoch": 5.425400226150092,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 22790
    },
    {
      "epoch": 5.425638278878772,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7483,
      "step": 22791
    },
    {
      "epoch": 5.425876331607451,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 22792
    },
    {
      "epoch": 5.42611438433613,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 22793
    },
    {
      "epoch": 5.42635243706481,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 22794
    },
    {
      "epoch": 5.426590489793489,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 22795
    },
    {
      "epoch": 5.426828542522169,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2158,
      "step": 22796
    },
    {
      "epoch": 5.427066595250848,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 22797
    },
    {
      "epoch": 5.427304647979527,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 22798
    },
    {
      "epoch": 5.427542700708207,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 22799
    },
    {
      "epoch": 5.4277807534368865,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 22800
    },
    {
      "epoch": 5.4277807534368865,
      "eval_runtime": 0.5284,
      "eval_samples_per_second": 1.892,
      "eval_steps_per_second": 1.892,
      "step": 22800
    },
    {
      "epoch": 5.428018806165566,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4802,
      "step": 22801
    },
    {
      "epoch": 5.428256858894245,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 22802
    },
    {
      "epoch": 5.428494911622924,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 22803
    },
    {
      "epoch": 5.428732964351604,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 22804
    },
    {
      "epoch": 5.428971017080284,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 22805
    },
    {
      "epoch": 5.429209069808962,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 22806
    },
    {
      "epoch": 5.429447122537642,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 22807
    },
    {
      "epoch": 5.4296851752663216,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 22808
    },
    {
      "epoch": 5.429923227995001,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 22809
    },
    {
      "epoch": 5.430161280723681,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.295,
      "step": 22810
    },
    {
      "epoch": 5.4303993334523595,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 22811
    },
    {
      "epoch": 5.430637386181039,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 22812
    },
    {
      "epoch": 5.430875438909719,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 22813
    },
    {
      "epoch": 5.431113491638398,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 22814
    },
    {
      "epoch": 5.431351544367077,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2812,
      "step": 22815
    },
    {
      "epoch": 5.431589597095757,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 22816
    },
    {
      "epoch": 5.431827649824436,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 22817
    },
    {
      "epoch": 5.432065702553116,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 22818
    },
    {
      "epoch": 5.4323037552817945,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 22819
    },
    {
      "epoch": 5.432541808010474,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4395,
      "step": 22820
    },
    {
      "epoch": 5.432779860739154,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 22821
    },
    {
      "epoch": 5.433017913467833,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3335,
      "step": 22822
    },
    {
      "epoch": 5.433255966196512,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 22823
    },
    {
      "epoch": 5.433494018925192,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 22824
    },
    {
      "epoch": 5.433732071653871,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 22825
    },
    {
      "epoch": 5.433970124382551,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 22826
    },
    {
      "epoch": 5.4342081771112305,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.5262,
      "step": 22827
    },
    {
      "epoch": 5.434446229839909,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 22828
    },
    {
      "epoch": 5.434684282568589,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 22829
    },
    {
      "epoch": 5.434922335297268,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 22830
    },
    {
      "epoch": 5.435160388025948,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 22831
    },
    {
      "epoch": 5.435398440754627,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 22832
    },
    {
      "epoch": 5.435636493483306,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 22833
    },
    {
      "epoch": 5.435874546211986,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 22834
    },
    {
      "epoch": 5.436112598940666,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 22835
    },
    {
      "epoch": 5.436350651669345,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 22836
    },
    {
      "epoch": 5.436588704398024,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 22837
    },
    {
      "epoch": 5.4368267571267035,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 22838
    },
    {
      "epoch": 5.437064809855383,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 22839
    },
    {
      "epoch": 5.437302862584063,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 22840
    },
    {
      "epoch": 5.437540915312741,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 22841
    },
    {
      "epoch": 5.437778968041421,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7056,
      "step": 22842
    },
    {
      "epoch": 5.438017020770101,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1813,
      "step": 22843
    },
    {
      "epoch": 5.43825507349878,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 22844
    },
    {
      "epoch": 5.43849312622746,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 22845
    },
    {
      "epoch": 5.438731178956139,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 22846
    },
    {
      "epoch": 5.438969231684818,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 22847
    },
    {
      "epoch": 5.439207284413498,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 22848
    },
    {
      "epoch": 5.439445337142177,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 22849
    },
    {
      "epoch": 5.439683389870856,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 22850
    },
    {
      "epoch": 5.439921442599536,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 22851
    },
    {
      "epoch": 5.440159495328215,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 22852
    },
    {
      "epoch": 5.440397548056895,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 22853
    },
    {
      "epoch": 5.440635600785574,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 22854
    },
    {
      "epoch": 5.440873653514253,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 22855
    },
    {
      "epoch": 5.441111706242933,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 22856
    },
    {
      "epoch": 5.441349758971612,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 22857
    },
    {
      "epoch": 5.441587811700291,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 22858
    },
    {
      "epoch": 5.441825864428971,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 22859
    },
    {
      "epoch": 5.44206391715765,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 22860
    },
    {
      "epoch": 5.44230196988633,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 22861
    },
    {
      "epoch": 5.44254002261501,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 22862
    },
    {
      "epoch": 5.442778075343688,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 22863
    },
    {
      "epoch": 5.443016128072368,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 22864
    },
    {
      "epoch": 5.4432541808010475,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 22865
    },
    {
      "epoch": 5.443492233529727,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 22866
    },
    {
      "epoch": 5.443730286258406,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 22867
    },
    {
      "epoch": 5.443968338987085,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3558,
      "step": 22868
    },
    {
      "epoch": 5.444206391715765,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 22869
    },
    {
      "epoch": 5.444444444444445,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 22870
    },
    {
      "epoch": 5.444682497173124,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 22871
    },
    {
      "epoch": 5.444920549901803,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 22872
    },
    {
      "epoch": 5.445158602630483,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 22873
    },
    {
      "epoch": 5.445396655359162,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 22874
    },
    {
      "epoch": 5.445634708087842,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 22875
    },
    {
      "epoch": 5.4458727608165205,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 22876
    },
    {
      "epoch": 5.4461108135452,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 22877
    },
    {
      "epoch": 5.44634886627388,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 22878
    },
    {
      "epoch": 5.446586919002559,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 22879
    },
    {
      "epoch": 5.446824971731239,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 22880
    },
    {
      "epoch": 5.447063024459918,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 22881
    },
    {
      "epoch": 5.447301077188597,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5922,
      "step": 22882
    },
    {
      "epoch": 5.447539129917277,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 22883
    },
    {
      "epoch": 5.4477771826459565,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 22884
    },
    {
      "epoch": 5.448015235374635,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 22885
    },
    {
      "epoch": 5.448253288103315,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 22886
    },
    {
      "epoch": 5.448491340831994,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 22887
    },
    {
      "epoch": 5.448729393560674,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 22888
    },
    {
      "epoch": 5.448967446289353,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 22889
    },
    {
      "epoch": 5.449205499018032,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 22890
    },
    {
      "epoch": 5.449443551746712,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 22891
    },
    {
      "epoch": 5.4496816044753915,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 22892
    },
    {
      "epoch": 5.44991965720407,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 22893
    },
    {
      "epoch": 5.45015770993275,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 22894
    },
    {
      "epoch": 5.450395762661429,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 22895
    },
    {
      "epoch": 5.450633815390109,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.331,
      "step": 22896
    },
    {
      "epoch": 5.450871868118789,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 22897
    },
    {
      "epoch": 5.451109920847467,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 22898
    },
    {
      "epoch": 5.451347973576147,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 22899
    },
    {
      "epoch": 5.451586026304827,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 22900
    },
    {
      "epoch": 5.451586026304827,
      "eval_runtime": 0.5374,
      "eval_samples_per_second": 1.861,
      "eval_steps_per_second": 1.861,
      "step": 22900
    },
    {
      "epoch": 5.451824079033506,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 22901
    },
    {
      "epoch": 5.452062131762185,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 22902
    },
    {
      "epoch": 5.4523001844908645,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4417,
      "step": 22903
    },
    {
      "epoch": 5.452538237219544,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 22904
    },
    {
      "epoch": 5.452776289948224,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 22905
    },
    {
      "epoch": 5.453014342676903,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 22906
    },
    {
      "epoch": 5.453252395405582,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 22907
    },
    {
      "epoch": 5.453490448134262,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 22908
    },
    {
      "epoch": 5.453728500862941,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 22909
    },
    {
      "epoch": 5.453966553591621,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.2695,
      "step": 22910
    },
    {
      "epoch": 5.4542046063203,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 22911
    },
    {
      "epoch": 5.454442659048979,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 22912
    },
    {
      "epoch": 5.454680711777659,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 22913
    },
    {
      "epoch": 5.454918764506338,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 22914
    },
    {
      "epoch": 5.455156817235018,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 22915
    },
    {
      "epoch": 5.455394869963697,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 22916
    },
    {
      "epoch": 5.455632922692376,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 22917
    },
    {
      "epoch": 5.455870975421056,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 22918
    },
    {
      "epoch": 5.4561090281497355,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3855,
      "step": 22919
    },
    {
      "epoch": 5.456347080878414,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 22920
    },
    {
      "epoch": 5.456585133607094,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 22921
    },
    {
      "epoch": 5.4568231863357735,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 22922
    },
    {
      "epoch": 5.457061239064453,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 22923
    },
    {
      "epoch": 5.457299291793132,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 22924
    },
    {
      "epoch": 5.457537344521811,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 22925
    },
    {
      "epoch": 5.457775397250491,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 22926
    },
    {
      "epoch": 5.458013449979171,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 22927
    },
    {
      "epoch": 5.45825150270785,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 22928
    },
    {
      "epoch": 5.458489555436529,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2507,
      "step": 22929
    },
    {
      "epoch": 5.4587276081652085,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 22930
    },
    {
      "epoch": 5.458965660893888,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 22931
    },
    {
      "epoch": 5.459203713622568,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 22932
    },
    {
      "epoch": 5.4594417663512465,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7248,
      "step": 22933
    },
    {
      "epoch": 5.459679819079926,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 22934
    },
    {
      "epoch": 5.459917871808606,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 22935
    },
    {
      "epoch": 5.460155924537285,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.4413,
      "step": 22936
    },
    {
      "epoch": 5.460393977265964,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 22937
    },
    {
      "epoch": 5.460632029994644,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 22938
    },
    {
      "epoch": 5.460870082723323,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 22939
    },
    {
      "epoch": 5.461108135452003,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 22940
    },
    {
      "epoch": 5.461346188180682,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 22941
    },
    {
      "epoch": 5.461584240909361,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5005,
      "step": 22942
    },
    {
      "epoch": 5.461822293638041,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 22943
    },
    {
      "epoch": 5.46206034636672,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 22944
    },
    {
      "epoch": 5.4622983990954,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3138,
      "step": 22945
    },
    {
      "epoch": 5.462536451824079,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 22946
    },
    {
      "epoch": 5.462774504552758,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 22947
    },
    {
      "epoch": 5.463012557281438,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 22948
    },
    {
      "epoch": 5.4632506100101175,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 22949
    },
    {
      "epoch": 5.463488662738797,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 22950
    },
    {
      "epoch": 5.463726715467476,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 22951
    },
    {
      "epoch": 5.463964768196155,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4401,
      "step": 22952
    },
    {
      "epoch": 5.464202820924835,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 22953
    },
    {
      "epoch": 5.464440873653515,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 22954
    },
    {
      "epoch": 5.464678926382193,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2507,
      "step": 22955
    },
    {
      "epoch": 5.464916979110873,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 22956
    },
    {
      "epoch": 5.4651550318395525,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2761,
      "step": 22957
    },
    {
      "epoch": 5.465393084568232,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 22958
    },
    {
      "epoch": 5.465631137296911,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 22959
    },
    {
      "epoch": 5.4658691900255905,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 22960
    },
    {
      "epoch": 5.46610724275427,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 22961
    },
    {
      "epoch": 5.46634529548295,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 22962
    },
    {
      "epoch": 5.466583348211629,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 22963
    },
    {
      "epoch": 5.466821400940308,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 22964
    },
    {
      "epoch": 5.467059453668988,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 22965
    },
    {
      "epoch": 5.467297506397667,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 22966
    },
    {
      "epoch": 5.467535559126347,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 22967
    },
    {
      "epoch": 5.4677736118550255,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 22968
    },
    {
      "epoch": 5.468011664583705,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 22969
    },
    {
      "epoch": 5.468249717312385,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 22970
    },
    {
      "epoch": 5.468487770041064,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 22971
    },
    {
      "epoch": 5.468725822769743,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 22972
    },
    {
      "epoch": 5.468963875498423,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 22973
    },
    {
      "epoch": 5.469201928227102,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 22974
    },
    {
      "epoch": 5.469439980955782,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 22975
    },
    {
      "epoch": 5.4696780336844615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9176,
      "step": 22976
    },
    {
      "epoch": 5.46991608641314,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 22977
    },
    {
      "epoch": 5.47015413914182,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 22978
    },
    {
      "epoch": 5.470392191870499,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3047,
      "step": 22979
    },
    {
      "epoch": 5.470630244599179,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1992,
      "step": 22980
    },
    {
      "epoch": 5.470868297327858,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 22981
    },
    {
      "epoch": 5.471106350056537,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 22982
    },
    {
      "epoch": 5.471344402785217,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 22983
    },
    {
      "epoch": 5.4715824555138965,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2962,
      "step": 22984
    },
    {
      "epoch": 5.471820508242576,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 22985
    },
    {
      "epoch": 5.472058560971255,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 22986
    },
    {
      "epoch": 5.4722966136999345,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 22987
    },
    {
      "epoch": 5.472534666428614,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 22988
    },
    {
      "epoch": 5.472772719157294,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 22989
    },
    {
      "epoch": 5.473010771885972,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 22990
    },
    {
      "epoch": 5.473248824614652,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 22991
    },
    {
      "epoch": 5.473486877343332,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3507,
      "step": 22992
    },
    {
      "epoch": 5.473724930072011,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 22993
    },
    {
      "epoch": 5.473962982800691,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 22994
    },
    {
      "epoch": 5.4742010355293695,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 22995
    },
    {
      "epoch": 5.474439088258049,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 22996
    },
    {
      "epoch": 5.474677140986729,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 22997
    },
    {
      "epoch": 5.474915193715408,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 22998
    },
    {
      "epoch": 5.475153246444087,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 22999
    },
    {
      "epoch": 5.475391299172767,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 23000
    },
    {
      "epoch": 5.475391299172767,
      "eval_runtime": 0.5146,
      "eval_samples_per_second": 1.943,
      "eval_steps_per_second": 1.943,
      "step": 23000
    },
    {
      "epoch": 5.475629351901446,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 23001
    },
    {
      "epoch": 5.475867404630126,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 23002
    },
    {
      "epoch": 5.476105457358805,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 23003
    },
    {
      "epoch": 5.476343510087484,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 23004
    },
    {
      "epoch": 5.476581562816164,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 23005
    },
    {
      "epoch": 5.476819615544843,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 23006
    },
    {
      "epoch": 5.477057668273522,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 23007
    },
    {
      "epoch": 5.477295721002202,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 23008
    },
    {
      "epoch": 5.477533773730881,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 23009
    },
    {
      "epoch": 5.477771826459561,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 23010
    },
    {
      "epoch": 5.478009879188241,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 23011
    },
    {
      "epoch": 5.478247931916919,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3896,
      "step": 23012
    },
    {
      "epoch": 5.478485984645599,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 23013
    },
    {
      "epoch": 5.4787240373742785,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 23014
    },
    {
      "epoch": 5.478962090102958,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 23015
    },
    {
      "epoch": 5.479200142831637,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 23016
    },
    {
      "epoch": 5.479438195560316,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 23017
    },
    {
      "epoch": 5.479676248288996,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 23018
    },
    {
      "epoch": 5.479914301017676,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 23019
    },
    {
      "epoch": 5.480152353746355,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 23020
    },
    {
      "epoch": 5.480390406475034,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 23021
    },
    {
      "epoch": 5.4806284592037136,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 23022
    },
    {
      "epoch": 5.480866511932393,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 23023
    },
    {
      "epoch": 5.481104564661073,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 23024
    },
    {
      "epoch": 5.4813426173897515,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 23025
    },
    {
      "epoch": 5.481580670118431,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 23026
    },
    {
      "epoch": 5.481818722847111,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 23027
    },
    {
      "epoch": 5.48205677557579,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 23028
    },
    {
      "epoch": 5.48229482830447,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 23029
    },
    {
      "epoch": 5.482532881033149,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 23030
    },
    {
      "epoch": 5.482770933761828,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 23031
    },
    {
      "epoch": 5.483008986490508,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5345,
      "step": 23032
    },
    {
      "epoch": 5.483247039219187,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 23033
    },
    {
      "epoch": 5.483485091947866,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 23034
    },
    {
      "epoch": 5.483723144676546,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 23035
    },
    {
      "epoch": 5.483961197405225,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 23036
    },
    {
      "epoch": 5.484199250133905,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7939,
      "step": 23037
    },
    {
      "epoch": 5.484437302862584,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 23038
    },
    {
      "epoch": 5.484675355591263,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 23039
    },
    {
      "epoch": 5.484913408319943,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 23040
    },
    {
      "epoch": 5.4851514610486225,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 23041
    },
    {
      "epoch": 5.485389513777301,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 23042
    },
    {
      "epoch": 5.485627566505981,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 23043
    },
    {
      "epoch": 5.48586561923466,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 23044
    },
    {
      "epoch": 5.48610367196334,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 23045
    },
    {
      "epoch": 5.48634172469202,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7122,
      "step": 23046
    },
    {
      "epoch": 5.486579777420698,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 23047
    },
    {
      "epoch": 5.486817830149378,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 23048
    },
    {
      "epoch": 5.487055882878058,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 23049
    },
    {
      "epoch": 5.487293935606737,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 23050
    },
    {
      "epoch": 5.487531988335416,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 23051
    },
    {
      "epoch": 5.4877700410640955,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 23052
    },
    {
      "epoch": 5.488008093792775,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 23053
    },
    {
      "epoch": 5.488246146521455,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 23054
    },
    {
      "epoch": 5.488484199250134,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 23055
    },
    {
      "epoch": 5.488722251978813,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 23056
    },
    {
      "epoch": 5.488960304707493,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 23057
    },
    {
      "epoch": 5.489198357436172,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 23058
    },
    {
      "epoch": 5.489436410164852,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 23059
    },
    {
      "epoch": 5.489674462893531,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 23060
    },
    {
      "epoch": 5.48991251562221,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 23061
    },
    {
      "epoch": 5.49015056835089,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 23062
    },
    {
      "epoch": 5.490388621079569,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 23063
    },
    {
      "epoch": 5.490626673808249,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 23064
    },
    {
      "epoch": 5.490864726536928,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 23065
    },
    {
      "epoch": 5.491102779265607,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 23066
    },
    {
      "epoch": 5.491340831994287,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 23067
    },
    {
      "epoch": 5.4915788847229665,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 23068
    },
    {
      "epoch": 5.491816937451645,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7826,
      "step": 23069
    },
    {
      "epoch": 5.492054990180325,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 23070
    },
    {
      "epoch": 5.492293042909004,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 23071
    },
    {
      "epoch": 5.492531095637684,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 23072
    },
    {
      "epoch": 5.492769148366363,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3377,
      "step": 23073
    },
    {
      "epoch": 5.493007201095042,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 23074
    },
    {
      "epoch": 5.493245253823722,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3172,
      "step": 23075
    },
    {
      "epoch": 5.493483306552402,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.3483,
      "step": 23076
    },
    {
      "epoch": 5.49372135928108,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 23077
    },
    {
      "epoch": 5.49395941200976,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 23078
    },
    {
      "epoch": 5.4941974647384395,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 23079
    },
    {
      "epoch": 5.494435517467119,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 23080
    },
    {
      "epoch": 5.494673570195799,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 23081
    },
    {
      "epoch": 5.494911622924477,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 23082
    },
    {
      "epoch": 5.495149675653157,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.5863,
      "step": 23083
    },
    {
      "epoch": 5.495387728381837,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3477,
      "step": 23084
    },
    {
      "epoch": 5.495625781110516,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 23085
    },
    {
      "epoch": 5.495863833839195,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 23086
    },
    {
      "epoch": 5.496101886567875,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 23087
    },
    {
      "epoch": 5.496339939296554,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 23088
    },
    {
      "epoch": 5.496577992025234,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 23089
    },
    {
      "epoch": 5.496816044753913,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 23090
    },
    {
      "epoch": 5.497054097482592,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 23091
    },
    {
      "epoch": 5.497292150211272,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 23092
    },
    {
      "epoch": 5.497530202939951,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 23093
    },
    {
      "epoch": 5.497768255668631,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3692,
      "step": 23094
    },
    {
      "epoch": 5.49800630839731,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 23095
    },
    {
      "epoch": 5.498244361125989,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 23096
    },
    {
      "epoch": 5.498482413854669,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 23097
    },
    {
      "epoch": 5.4987204665833485,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 23098
    },
    {
      "epoch": 5.498958519312028,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 23099
    },
    {
      "epoch": 5.499196572040707,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 23100
    },
    {
      "epoch": 5.499196572040707,
      "eval_runtime": 0.5411,
      "eval_samples_per_second": 1.848,
      "eval_steps_per_second": 1.848,
      "step": 23100
    },
    {
      "epoch": 5.499434624769386,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 23101
    },
    {
      "epoch": 5.499672677498066,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 23102
    },
    {
      "epoch": 5.499910730226746,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 23103
    },
    {
      "epoch": 5.500148782955424,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8496,
      "step": 23104
    },
    {
      "epoch": 5.500386835684104,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4348,
      "step": 23105
    },
    {
      "epoch": 5.5006248884127835,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 23106
    },
    {
      "epoch": 5.500862941141463,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 23107
    },
    {
      "epoch": 5.501100993870143,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 23108
    },
    {
      "epoch": 5.5013390465988214,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3433,
      "step": 23109
    },
    {
      "epoch": 5.501577099327501,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 23110
    },
    {
      "epoch": 5.501815152056181,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3098,
      "step": 23111
    },
    {
      "epoch": 5.502053204784859,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 23112
    },
    {
      "epoch": 5.502291257513539,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 23113
    },
    {
      "epoch": 5.502529310242219,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 23114
    },
    {
      "epoch": 5.502767362970898,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 23115
    },
    {
      "epoch": 5.503005415699578,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 23116
    },
    {
      "epoch": 5.5032434684282565,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 23117
    },
    {
      "epoch": 5.503481521156936,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 23118
    },
    {
      "epoch": 5.503719573885616,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 23119
    },
    {
      "epoch": 5.503957626614295,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 23120
    },
    {
      "epoch": 5.504195679342974,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 23121
    },
    {
      "epoch": 5.504433732071654,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9533,
      "step": 23122
    },
    {
      "epoch": 5.504671784800333,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3166,
      "step": 23123
    },
    {
      "epoch": 5.504909837529013,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 23124
    },
    {
      "epoch": 5.5051478902576925,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 23125
    },
    {
      "epoch": 5.505385942986371,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 23126
    },
    {
      "epoch": 5.505623995715051,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 23127
    },
    {
      "epoch": 5.50586204844373,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 23128
    },
    {
      "epoch": 5.50610010117241,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 23129
    },
    {
      "epoch": 5.506338153901089,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 23130
    },
    {
      "epoch": 5.506576206629768,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 23131
    },
    {
      "epoch": 5.506814259358448,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 23132
    },
    {
      "epoch": 5.5070523120871275,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 23133
    },
    {
      "epoch": 5.507290364815807,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 23134
    },
    {
      "epoch": 5.507528417544486,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 23135
    },
    {
      "epoch": 5.5077664702731655,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 23136
    },
    {
      "epoch": 5.508004523001845,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 23137
    },
    {
      "epoch": 5.508242575730525,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 23138
    },
    {
      "epoch": 5.508480628459203,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 23139
    },
    {
      "epoch": 5.508718681187883,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 23140
    },
    {
      "epoch": 5.508956733916563,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 23141
    },
    {
      "epoch": 5.509194786645242,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 23142
    },
    {
      "epoch": 5.509432839373922,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 23143
    },
    {
      "epoch": 5.5096708921026005,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 23144
    },
    {
      "epoch": 5.50990894483128,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 23145
    },
    {
      "epoch": 5.51014699755996,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3662,
      "step": 23146
    },
    {
      "epoch": 5.5103850502886385,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 23147
    },
    {
      "epoch": 5.510623103017318,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 23148
    },
    {
      "epoch": 5.510861155745998,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 23149
    },
    {
      "epoch": 5.511099208474677,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 23150
    },
    {
      "epoch": 5.511337261203357,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 23151
    },
    {
      "epoch": 5.511575313932036,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 23152
    },
    {
      "epoch": 5.511813366660715,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 23153
    },
    {
      "epoch": 5.512051419389395,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2979,
      "step": 23154
    },
    {
      "epoch": 5.512289472118074,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 23155
    },
    {
      "epoch": 5.512527524846753,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 23156
    },
    {
      "epoch": 5.512765577575433,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 23157
    },
    {
      "epoch": 5.513003630304112,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 23158
    },
    {
      "epoch": 5.513241683032792,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 23159
    },
    {
      "epoch": 5.5134797357614715,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 23160
    },
    {
      "epoch": 5.51371778849015,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 23161
    },
    {
      "epoch": 5.51395584121883,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 23162
    },
    {
      "epoch": 5.5141938939475095,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 23163
    },
    {
      "epoch": 5.514431946676189,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4138,
      "step": 23164
    },
    {
      "epoch": 5.514669999404868,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 23165
    },
    {
      "epoch": 5.514908052133547,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 23166
    },
    {
      "epoch": 5.515146104862227,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 23167
    },
    {
      "epoch": 5.515384157590907,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 23168
    },
    {
      "epoch": 5.515622210319586,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 23169
    },
    {
      "epoch": 5.515860263048265,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 23170
    },
    {
      "epoch": 5.5160983157769445,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 23171
    },
    {
      "epoch": 5.516336368505624,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 23172
    },
    {
      "epoch": 5.516574421234304,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 23173
    },
    {
      "epoch": 5.5168124739629825,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 23174
    },
    {
      "epoch": 5.517050526691662,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 23175
    },
    {
      "epoch": 5.517288579420342,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 23176
    },
    {
      "epoch": 5.517526632149021,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 23177
    },
    {
      "epoch": 5.517764684877701,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 23178
    },
    {
      "epoch": 5.51800273760638,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 23179
    },
    {
      "epoch": 5.518240790335059,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 23180
    },
    {
      "epoch": 5.518478843063739,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 23181
    },
    {
      "epoch": 5.5187168957924175,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 23182
    },
    {
      "epoch": 5.518954948521097,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 23183
    },
    {
      "epoch": 5.519193001249777,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 23184
    },
    {
      "epoch": 5.519431053978456,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 23185
    },
    {
      "epoch": 5.519669106707136,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 23186
    },
    {
      "epoch": 5.519907159435815,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 23187
    },
    {
      "epoch": 5.520145212164494,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 23188
    },
    {
      "epoch": 5.520383264893174,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 23189
    },
    {
      "epoch": 5.5206213176218535,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9304,
      "step": 23190
    },
    {
      "epoch": 5.520859370350532,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4359,
      "step": 23191
    },
    {
      "epoch": 5.521097423079212,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 23192
    },
    {
      "epoch": 5.521335475807891,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 23193
    },
    {
      "epoch": 5.521573528536571,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 23194
    },
    {
      "epoch": 5.521811581265251,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3607,
      "step": 23195
    },
    {
      "epoch": 5.522049633993929,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 23196
    },
    {
      "epoch": 5.522287686722609,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 23197
    },
    {
      "epoch": 5.5225257394512886,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 23198
    },
    {
      "epoch": 5.522763792179968,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 23199
    },
    {
      "epoch": 5.523001844908647,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 23200
    },
    {
      "epoch": 5.523001844908647,
      "eval_runtime": 0.5178,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 23200
    },
    {
      "epoch": 5.5232398976373265,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 23201
    },
    {
      "epoch": 5.523477950366006,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 23202
    },
    {
      "epoch": 5.523716003094686,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 23203
    },
    {
      "epoch": 5.523954055823365,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 23204
    },
    {
      "epoch": 5.524192108552044,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 23205
    },
    {
      "epoch": 5.524430161280724,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 23206
    },
    {
      "epoch": 5.524668214009403,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 23207
    },
    {
      "epoch": 5.524906266738083,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 23208
    },
    {
      "epoch": 5.5251443194667615,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 23209
    },
    {
      "epoch": 5.525382372195441,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 23210
    },
    {
      "epoch": 5.525620424924121,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 23211
    },
    {
      "epoch": 5.5258584776528,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 23212
    },
    {
      "epoch": 5.52609653038148,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 23213
    },
    {
      "epoch": 5.526334583110159,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 23214
    },
    {
      "epoch": 5.526572635838838,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 23215
    },
    {
      "epoch": 5.526810688567518,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 23216
    },
    {
      "epoch": 5.527048741296197,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 23217
    },
    {
      "epoch": 5.527286794024876,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 23218
    },
    {
      "epoch": 5.527524846753556,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 23219
    },
    {
      "epoch": 5.527762899482235,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 23220
    },
    {
      "epoch": 5.528000952210915,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 23221
    },
    {
      "epoch": 5.528239004939594,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 23222
    },
    {
      "epoch": 5.528477057668273,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 23223
    },
    {
      "epoch": 5.528715110396953,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 23224
    },
    {
      "epoch": 5.528953163125633,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 23225
    },
    {
      "epoch": 5.529191215854311,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.4064,
      "step": 23226
    },
    {
      "epoch": 5.529429268582991,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 23227
    },
    {
      "epoch": 5.5296673213116705,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 23228
    },
    {
      "epoch": 5.52990537404035,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 23229
    },
    {
      "epoch": 5.53014342676903,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2671,
      "step": 23230
    },
    {
      "epoch": 5.530381479497708,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 23231
    },
    {
      "epoch": 5.530619532226388,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 23232
    },
    {
      "epoch": 5.530857584955068,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 23233
    },
    {
      "epoch": 5.531095637683747,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 23234
    },
    {
      "epoch": 5.531333690412426,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.4154,
      "step": 23235
    },
    {
      "epoch": 5.531571743141106,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 23236
    },
    {
      "epoch": 5.531809795869785,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 23237
    },
    {
      "epoch": 5.532047848598465,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 23238
    },
    {
      "epoch": 5.532285901327144,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 23239
    },
    {
      "epoch": 5.532523954055823,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 23240
    },
    {
      "epoch": 5.532762006784503,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 23241
    },
    {
      "epoch": 5.533000059513182,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 23242
    },
    {
      "epoch": 5.533238112241862,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 23243
    },
    {
      "epoch": 5.533476164970541,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 23244
    },
    {
      "epoch": 5.53371421769922,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 23245
    },
    {
      "epoch": 5.5339522704279,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 23246
    },
    {
      "epoch": 5.534190323156579,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 23247
    },
    {
      "epoch": 5.534428375885259,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 23248
    },
    {
      "epoch": 5.534666428613938,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 23249
    },
    {
      "epoch": 5.534904481342617,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 23250
    },
    {
      "epoch": 5.535142534071297,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 23251
    },
    {
      "epoch": 5.535380586799977,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 23252
    },
    {
      "epoch": 5.535618639528655,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3203,
      "step": 23253
    },
    {
      "epoch": 5.535856692257335,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 23254
    },
    {
      "epoch": 5.5360947449860145,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 23255
    },
    {
      "epoch": 5.536332797714694,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 23256
    },
    {
      "epoch": 5.536570850443373,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 23257
    },
    {
      "epoch": 5.536808903172052,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 23258
    },
    {
      "epoch": 5.537046955900732,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 23259
    },
    {
      "epoch": 5.537285008629412,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 23260
    },
    {
      "epoch": 5.53752306135809,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7506,
      "step": 23261
    },
    {
      "epoch": 5.53776111408677,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 23262
    },
    {
      "epoch": 5.53799916681545,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 23263
    },
    {
      "epoch": 5.538237219544129,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2805,
      "step": 23264
    },
    {
      "epoch": 5.538475272272809,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 23265
    },
    {
      "epoch": 5.5387133250014875,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 23266
    },
    {
      "epoch": 5.538951377730167,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 23267
    },
    {
      "epoch": 5.539189430458847,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 23268
    },
    {
      "epoch": 5.539427483187526,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 23269
    },
    {
      "epoch": 5.539665535916205,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 23270
    },
    {
      "epoch": 5.539903588644885,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 23271
    },
    {
      "epoch": 5.540141641373564,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 23272
    },
    {
      "epoch": 5.540379694102244,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 23273
    },
    {
      "epoch": 5.5406177468309235,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 23274
    },
    {
      "epoch": 5.540855799559602,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.4359,
      "step": 23275
    },
    {
      "epoch": 5.541093852288282,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 23276
    },
    {
      "epoch": 5.541331905016961,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 23277
    },
    {
      "epoch": 5.541569957745641,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 23278
    },
    {
      "epoch": 5.54180801047432,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 23279
    },
    {
      "epoch": 5.542046063202999,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3416,
      "step": 23280
    },
    {
      "epoch": 5.542284115931679,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8708,
      "step": 23281
    },
    {
      "epoch": 5.5425221686603585,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 23282
    },
    {
      "epoch": 5.542760221389038,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 23283
    },
    {
      "epoch": 5.542998274117717,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5304,
      "step": 23284
    },
    {
      "epoch": 5.5432363268463964,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 23285
    },
    {
      "epoch": 5.543474379575076,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 23286
    },
    {
      "epoch": 5.543712432303756,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 23287
    },
    {
      "epoch": 5.543950485032434,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 23288
    },
    {
      "epoch": 5.544188537761114,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 23289
    },
    {
      "epoch": 5.544426590489794,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 23290
    },
    {
      "epoch": 5.544664643218473,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 23291
    },
    {
      "epoch": 5.544902695947153,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3915,
      "step": 23292
    },
    {
      "epoch": 5.5451407486758315,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.84,
      "step": 23293
    },
    {
      "epoch": 5.545378801404511,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 23294
    },
    {
      "epoch": 5.545616854133191,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 23295
    },
    {
      "epoch": 5.545854906861869,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 23296
    },
    {
      "epoch": 5.546092959590549,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 23297
    },
    {
      "epoch": 5.546331012319229,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3701,
      "step": 23298
    },
    {
      "epoch": 5.546569065047908,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 23299
    },
    {
      "epoch": 5.546807117776588,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 23300
    },
    {
      "epoch": 5.546807117776588,
      "eval_runtime": 0.5247,
      "eval_samples_per_second": 1.906,
      "eval_steps_per_second": 1.906,
      "step": 23300
    },
    {
      "epoch": 5.547045170505267,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 23301
    },
    {
      "epoch": 5.547283223233946,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 23302
    },
    {
      "epoch": 5.547521275962626,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 23303
    },
    {
      "epoch": 5.547759328691305,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 23304
    },
    {
      "epoch": 5.547997381419984,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 23305
    },
    {
      "epoch": 5.548235434148664,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 23306
    },
    {
      "epoch": 5.548473486877343,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 23307
    },
    {
      "epoch": 5.548711539606023,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 23308
    },
    {
      "epoch": 5.5489495923347025,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 23309
    },
    {
      "epoch": 5.549187645063381,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 23310
    },
    {
      "epoch": 5.549425697792061,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 23311
    },
    {
      "epoch": 5.5496637505207405,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 23312
    },
    {
      "epoch": 5.54990180324942,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4294,
      "step": 23313
    },
    {
      "epoch": 5.550139855978099,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 23314
    },
    {
      "epoch": 5.550377908706778,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 23315
    },
    {
      "epoch": 5.550615961435458,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 23316
    },
    {
      "epoch": 5.550854014164138,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 23317
    },
    {
      "epoch": 5.551092066892817,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 23318
    },
    {
      "epoch": 5.551330119621496,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 23319
    },
    {
      "epoch": 5.5515681723501755,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 23320
    },
    {
      "epoch": 5.551806225078855,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 23321
    },
    {
      "epoch": 5.552044277807535,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 23322
    },
    {
      "epoch": 5.5522823305362135,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4515,
      "step": 23323
    },
    {
      "epoch": 5.552520383264893,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 23324
    },
    {
      "epoch": 5.552758435993573,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 23325
    },
    {
      "epoch": 5.552996488722252,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 23326
    },
    {
      "epoch": 5.553234541450932,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2762,
      "step": 23327
    },
    {
      "epoch": 5.553472594179611,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 23328
    },
    {
      "epoch": 5.55371064690829,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 23329
    },
    {
      "epoch": 5.55394869963697,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 23330
    },
    {
      "epoch": 5.5541867523656485,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 23331
    },
    {
      "epoch": 5.554424805094328,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 23332
    },
    {
      "epoch": 5.554662857823008,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 23333
    },
    {
      "epoch": 5.554900910551687,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 23334
    },
    {
      "epoch": 5.555138963280367,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 23335
    },
    {
      "epoch": 5.555377016009046,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 23336
    },
    {
      "epoch": 5.555615068737725,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 23337
    },
    {
      "epoch": 5.555853121466405,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 23338
    },
    {
      "epoch": 5.5560911741950845,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 23339
    },
    {
      "epoch": 5.556329226923763,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 23340
    },
    {
      "epoch": 5.556567279652443,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 23341
    },
    {
      "epoch": 5.556805332381122,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 23342
    },
    {
      "epoch": 5.557043385109802,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 23343
    },
    {
      "epoch": 5.557281437838482,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 23344
    },
    {
      "epoch": 5.55751949056716,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 23345
    },
    {
      "epoch": 5.55775754329584,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 23346
    },
    {
      "epoch": 5.5579955960245195,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 23347
    },
    {
      "epoch": 5.558233648753199,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 23348
    },
    {
      "epoch": 5.558471701481878,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 23349
    },
    {
      "epoch": 5.5587097542105575,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 23350
    },
    {
      "epoch": 5.558947806939237,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 23351
    },
    {
      "epoch": 5.559185859667917,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 23352
    },
    {
      "epoch": 5.559423912396596,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 23353
    },
    {
      "epoch": 5.559661965125275,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 23354
    },
    {
      "epoch": 5.559900017853955,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 23355
    },
    {
      "epoch": 5.560138070582634,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4127,
      "step": 23356
    },
    {
      "epoch": 5.560376123311314,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 23357
    },
    {
      "epoch": 5.5606141760399925,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 23358
    },
    {
      "epoch": 5.560852228768672,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 23359
    },
    {
      "epoch": 5.561090281497352,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 23360
    },
    {
      "epoch": 5.561328334226031,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 23361
    },
    {
      "epoch": 5.561566386954711,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 23362
    },
    {
      "epoch": 5.56180443968339,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 23363
    },
    {
      "epoch": 5.562042492412069,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8381,
      "step": 23364
    },
    {
      "epoch": 5.562280545140749,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 23365
    },
    {
      "epoch": 5.562518597869428,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 23366
    },
    {
      "epoch": 5.562756650598107,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 23367
    },
    {
      "epoch": 5.562994703326787,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4365,
      "step": 23368
    },
    {
      "epoch": 5.563232756055466,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 23369
    },
    {
      "epoch": 5.563470808784146,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 23370
    },
    {
      "epoch": 5.563708861512825,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 23371
    },
    {
      "epoch": 5.563946914241504,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 23372
    },
    {
      "epoch": 5.564184966970184,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 23373
    },
    {
      "epoch": 5.5644230196988635,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 23374
    },
    {
      "epoch": 5.564661072427542,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 23375
    },
    {
      "epoch": 5.564899125156222,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 23376
    },
    {
      "epoch": 5.5651371778849015,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 23377
    },
    {
      "epoch": 5.565375230613581,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 23378
    },
    {
      "epoch": 5.565613283342261,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 23379
    },
    {
      "epoch": 5.565851336070939,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 23380
    },
    {
      "epoch": 5.566089388799619,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 23381
    },
    {
      "epoch": 5.566327441528299,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 23382
    },
    {
      "epoch": 5.566565494256978,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 23383
    },
    {
      "epoch": 5.566803546985657,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 23384
    },
    {
      "epoch": 5.5670415997143365,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 23385
    },
    {
      "epoch": 5.567279652443016,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 23386
    },
    {
      "epoch": 5.567517705171696,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9357,
      "step": 23387
    },
    {
      "epoch": 5.567755757900375,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 23388
    },
    {
      "epoch": 5.567993810629054,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 23389
    },
    {
      "epoch": 5.568231863357734,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 23390
    },
    {
      "epoch": 5.568469916086413,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 23391
    },
    {
      "epoch": 5.568707968815093,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4603,
      "step": 23392
    },
    {
      "epoch": 5.568946021543772,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8532,
      "step": 23393
    },
    {
      "epoch": 5.569184074272451,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 23394
    },
    {
      "epoch": 5.569422127001131,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.764,
      "step": 23395
    },
    {
      "epoch": 5.56966017972981,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 23396
    },
    {
      "epoch": 5.56989823245849,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2709,
      "step": 23397
    },
    {
      "epoch": 5.570136285187169,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 23398
    },
    {
      "epoch": 5.570374337915848,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 23399
    },
    {
      "epoch": 5.570612390644528,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 23400
    },
    {
      "epoch": 5.570612390644528,
      "eval_runtime": 0.5191,
      "eval_samples_per_second": 1.926,
      "eval_steps_per_second": 1.926,
      "step": 23400
    },
    {
      "epoch": 5.570850443373207,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 23401
    },
    {
      "epoch": 5.571088496101886,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 23402
    },
    {
      "epoch": 5.571326548830566,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 23403
    },
    {
      "epoch": 5.5715646015592455,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 23404
    },
    {
      "epoch": 5.571802654287925,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 23405
    },
    {
      "epoch": 5.572040707016604,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2916,
      "step": 23406
    },
    {
      "epoch": 5.572278759745283,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5883,
      "step": 23407
    },
    {
      "epoch": 5.572516812473963,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 23408
    },
    {
      "epoch": 5.572754865202643,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 23409
    },
    {
      "epoch": 5.572992917931321,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 23410
    },
    {
      "epoch": 5.573230970660001,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 23411
    },
    {
      "epoch": 5.5734690233886806,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 23412
    },
    {
      "epoch": 5.57370707611736,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 23413
    },
    {
      "epoch": 5.57394512884604,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 23414
    },
    {
      "epoch": 5.5741831815747185,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.745,
      "step": 23415
    },
    {
      "epoch": 5.574421234303398,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 23416
    },
    {
      "epoch": 5.574659287032078,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3127,
      "step": 23417
    },
    {
      "epoch": 5.574897339760757,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8675,
      "step": 23418
    },
    {
      "epoch": 5.575135392489436,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 23419
    },
    {
      "epoch": 5.575373445218116,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 23420
    },
    {
      "epoch": 5.575611497946795,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 23421
    },
    {
      "epoch": 5.575849550675475,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 23422
    },
    {
      "epoch": 5.576087603404154,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2202,
      "step": 23423
    },
    {
      "epoch": 5.576325656132833,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 23424
    },
    {
      "epoch": 5.576563708861513,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 23425
    },
    {
      "epoch": 5.576801761590192,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 23426
    },
    {
      "epoch": 5.577039814318872,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 23427
    },
    {
      "epoch": 5.577277867047551,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 23428
    },
    {
      "epoch": 5.57751591977623,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 23429
    },
    {
      "epoch": 5.57775397250491,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 23430
    },
    {
      "epoch": 5.5779920252335895,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 23431
    },
    {
      "epoch": 5.578230077962269,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 23432
    },
    {
      "epoch": 5.578468130690948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 23433
    },
    {
      "epoch": 5.578706183419627,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 23434
    },
    {
      "epoch": 5.578944236148307,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 23435
    },
    {
      "epoch": 5.579182288876987,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 23436
    },
    {
      "epoch": 5.579420341605665,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 23437
    },
    {
      "epoch": 5.579658394334345,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 23438
    },
    {
      "epoch": 5.579896447063025,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 23439
    },
    {
      "epoch": 5.580134499791704,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 23440
    },
    {
      "epoch": 5.580372552520383,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 23441
    },
    {
      "epoch": 5.5806106052490625,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 23442
    },
    {
      "epoch": 5.580848657977742,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 23443
    },
    {
      "epoch": 5.581086710706422,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 23444
    },
    {
      "epoch": 5.5813247634351,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 23445
    },
    {
      "epoch": 5.58156281616378,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 23446
    },
    {
      "epoch": 5.58180086889246,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7846,
      "step": 23447
    },
    {
      "epoch": 5.582038921621139,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 23448
    },
    {
      "epoch": 5.582276974349819,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 23449
    },
    {
      "epoch": 5.582515027078498,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 23450
    },
    {
      "epoch": 5.582753079807177,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 23451
    },
    {
      "epoch": 5.582991132535857,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 23452
    },
    {
      "epoch": 5.583229185264536,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 23453
    },
    {
      "epoch": 5.583467237993215,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 23454
    },
    {
      "epoch": 5.583705290721895,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 23455
    },
    {
      "epoch": 5.583943343450574,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 23456
    },
    {
      "epoch": 5.584181396179254,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 23457
    },
    {
      "epoch": 5.5844194489079335,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 23458
    },
    {
      "epoch": 5.584657501636612,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 23459
    },
    {
      "epoch": 5.584895554365292,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 23460
    },
    {
      "epoch": 5.585133607093971,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 23461
    },
    {
      "epoch": 5.585371659822651,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1517,
      "step": 23462
    },
    {
      "epoch": 5.58560971255133,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 23463
    },
    {
      "epoch": 5.585847765280009,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.766,
      "step": 23464
    },
    {
      "epoch": 5.586085818008689,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 23465
    },
    {
      "epoch": 5.586323870737369,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3576,
      "step": 23466
    },
    {
      "epoch": 5.586561923466048,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 23467
    },
    {
      "epoch": 5.586799976194727,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 23468
    },
    {
      "epoch": 5.5870380289234065,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 23469
    },
    {
      "epoch": 5.587276081652086,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 23470
    },
    {
      "epoch": 5.587514134380766,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 23471
    },
    {
      "epoch": 5.587752187109444,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 23472
    },
    {
      "epoch": 5.587990239838124,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 23473
    },
    {
      "epoch": 5.588228292566804,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 23474
    },
    {
      "epoch": 5.588466345295483,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 23475
    },
    {
      "epoch": 5.588704398024162,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.238,
      "step": 23476
    },
    {
      "epoch": 5.588942450752842,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 23477
    },
    {
      "epoch": 5.589180503481521,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 23478
    },
    {
      "epoch": 5.589418556210201,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 23479
    },
    {
      "epoch": 5.5896566089388795,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 23480
    },
    {
      "epoch": 5.589894661667559,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 23481
    },
    {
      "epoch": 5.590132714396239,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 23482
    },
    {
      "epoch": 5.590370767124918,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 23483
    },
    {
      "epoch": 5.590608819853598,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 23484
    },
    {
      "epoch": 5.590846872582277,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 23485
    },
    {
      "epoch": 5.591084925310956,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3511,
      "step": 23486
    },
    {
      "epoch": 5.591322978039636,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 23487
    },
    {
      "epoch": 5.5915610307683155,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 23488
    },
    {
      "epoch": 5.591799083496994,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 23489
    },
    {
      "epoch": 5.592037136225674,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 23490
    },
    {
      "epoch": 5.592275188954353,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.298,
      "step": 23491
    },
    {
      "epoch": 5.592513241683033,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4324,
      "step": 23492
    },
    {
      "epoch": 5.592751294411713,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 23493
    },
    {
      "epoch": 5.592989347140391,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 23494
    },
    {
      "epoch": 5.593227399869071,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 23495
    },
    {
      "epoch": 5.5934654525977505,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 23496
    },
    {
      "epoch": 5.59370350532643,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 23497
    },
    {
      "epoch": 5.593941558055109,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 23498
    },
    {
      "epoch": 5.5941796107837884,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 23499
    },
    {
      "epoch": 5.594417663512468,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3675,
      "step": 23500
    },
    {
      "epoch": 5.594417663512468,
      "eval_runtime": 0.5358,
      "eval_samples_per_second": 1.866,
      "eval_steps_per_second": 1.866,
      "step": 23500
    },
    {
      "epoch": 5.594655716241148,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 23501
    },
    {
      "epoch": 5.594893768969827,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 23502
    },
    {
      "epoch": 5.595131821698506,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 23503
    },
    {
      "epoch": 5.595369874427186,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1537,
      "step": 23504
    },
    {
      "epoch": 5.595607927155865,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 23505
    },
    {
      "epoch": 5.595845979884545,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 23506
    },
    {
      "epoch": 5.5960840326132235,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 23507
    },
    {
      "epoch": 5.596322085341903,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 23508
    },
    {
      "epoch": 5.596560138070583,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 23509
    },
    {
      "epoch": 5.596798190799262,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2008,
      "step": 23510
    },
    {
      "epoch": 5.597036243527942,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 23511
    },
    {
      "epoch": 5.597274296256621,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 23512
    },
    {
      "epoch": 5.5975123489853,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 23513
    },
    {
      "epoch": 5.59775040171398,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5487,
      "step": 23514
    },
    {
      "epoch": 5.597988454442659,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 23515
    },
    {
      "epoch": 5.598226507171338,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 23516
    },
    {
      "epoch": 5.598464559900018,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 23517
    },
    {
      "epoch": 5.598702612628697,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 23518
    },
    {
      "epoch": 5.598940665357377,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 23519
    },
    {
      "epoch": 5.599178718086056,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 23520
    },
    {
      "epoch": 5.599416770814735,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 23521
    },
    {
      "epoch": 5.599654823543415,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 23522
    },
    {
      "epoch": 5.5998928762720945,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 23523
    },
    {
      "epoch": 5.600130929000773,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 23524
    },
    {
      "epoch": 5.600368981729453,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 23525
    },
    {
      "epoch": 5.6006070344581325,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 23526
    },
    {
      "epoch": 5.600845087186812,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 23527
    },
    {
      "epoch": 5.601083139915492,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 23528
    },
    {
      "epoch": 5.60132119264417,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 23529
    },
    {
      "epoch": 5.60155924537285,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 23530
    },
    {
      "epoch": 5.60179729810153,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 23531
    },
    {
      "epoch": 5.602035350830209,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 23532
    },
    {
      "epoch": 5.602273403558888,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 23533
    },
    {
      "epoch": 5.6025114562875675,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 23534
    },
    {
      "epoch": 5.602749509016247,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 23535
    },
    {
      "epoch": 5.602987561744927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 23536
    },
    {
      "epoch": 5.603225614473606,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2157,
      "step": 23537
    },
    {
      "epoch": 5.603463667202285,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 23538
    },
    {
      "epoch": 5.603701719930965,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 23539
    },
    {
      "epoch": 5.603939772659644,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 23540
    },
    {
      "epoch": 5.604177825388324,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 23541
    },
    {
      "epoch": 5.604415878117003,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3726,
      "step": 23542
    },
    {
      "epoch": 5.604653930845682,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 23543
    },
    {
      "epoch": 5.604891983574362,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 23544
    },
    {
      "epoch": 5.605130036303041,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 23545
    },
    {
      "epoch": 5.605368089031721,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3718,
      "step": 23546
    },
    {
      "epoch": 5.6056061417604,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 23547
    },
    {
      "epoch": 5.605844194489079,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 23548
    },
    {
      "epoch": 5.606082247217759,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 23549
    },
    {
      "epoch": 5.606320299946438,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 23550
    },
    {
      "epoch": 5.606558352675117,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3914,
      "step": 23551
    },
    {
      "epoch": 5.606796405403797,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 23552
    },
    {
      "epoch": 5.6070344581324765,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 23553
    },
    {
      "epoch": 5.607272510861156,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 23554
    },
    {
      "epoch": 5.607510563589835,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 23555
    },
    {
      "epoch": 5.607748616318514,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 23556
    },
    {
      "epoch": 5.607986669047194,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 23557
    },
    {
      "epoch": 5.608224721775874,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 23558
    },
    {
      "epoch": 5.608462774504552,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 23559
    },
    {
      "epoch": 5.608700827233232,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 23560
    },
    {
      "epoch": 5.6089388799619115,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 23561
    },
    {
      "epoch": 5.609176932690591,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 23562
    },
    {
      "epoch": 5.609414985419271,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2694,
      "step": 23563
    },
    {
      "epoch": 5.6096530381479495,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 23564
    },
    {
      "epoch": 5.609891090876629,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2697,
      "step": 23565
    },
    {
      "epoch": 5.610129143605309,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.565,
      "step": 23566
    },
    {
      "epoch": 5.610367196333988,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4239,
      "step": 23567
    },
    {
      "epoch": 5.610605249062667,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 23568
    },
    {
      "epoch": 5.610843301791347,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 23569
    },
    {
      "epoch": 5.611081354520026,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 23570
    },
    {
      "epoch": 5.611319407248706,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 23571
    },
    {
      "epoch": 5.611557459977385,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 23572
    },
    {
      "epoch": 5.611795512706064,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 23573
    },
    {
      "epoch": 5.612033565434744,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 23574
    },
    {
      "epoch": 5.612271618163423,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 23575
    },
    {
      "epoch": 5.612509670892103,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 23576
    },
    {
      "epoch": 5.612747723620782,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.5339,
      "step": 23577
    },
    {
      "epoch": 5.612985776349461,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 23578
    },
    {
      "epoch": 5.613223829078141,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2043,
      "step": 23579
    },
    {
      "epoch": 5.6134618818068205,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2351,
      "step": 23580
    },
    {
      "epoch": 5.6136999345355,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 23581
    },
    {
      "epoch": 5.613937987264179,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 23582
    },
    {
      "epoch": 5.614176039992858,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 23583
    },
    {
      "epoch": 5.614414092721538,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 23584
    },
    {
      "epoch": 5.614652145450217,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 23585
    },
    {
      "epoch": 5.614890198178896,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4209,
      "step": 23586
    },
    {
      "epoch": 5.615128250907576,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 23587
    },
    {
      "epoch": 5.6153663036362556,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 23588
    },
    {
      "epoch": 5.615604356364935,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 23589
    },
    {
      "epoch": 5.615842409093614,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 23590
    },
    {
      "epoch": 5.6160804618222935,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 23591
    },
    {
      "epoch": 5.616318514550973,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 23592
    },
    {
      "epoch": 5.616556567279653,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 23593
    },
    {
      "epoch": 5.616794620008331,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 23594
    },
    {
      "epoch": 5.617032672737011,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 23595
    },
    {
      "epoch": 5.617270725465691,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3577,
      "step": 23596
    },
    {
      "epoch": 5.61750877819437,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 23597
    },
    {
      "epoch": 5.61774683092305,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 23598
    },
    {
      "epoch": 5.6179848836517285,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2116,
      "step": 23599
    },
    {
      "epoch": 5.618222936380408,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 23600
    },
    {
      "epoch": 5.618222936380408,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 23600
    },
    {
      "epoch": 5.618460989109088,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 23601
    },
    {
      "epoch": 5.618699041837767,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8523,
      "step": 23602
    },
    {
      "epoch": 5.618937094566446,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3132,
      "step": 23603
    },
    {
      "epoch": 5.619175147295126,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3267,
      "step": 23604
    },
    {
      "epoch": 5.619413200023805,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 23605
    },
    {
      "epoch": 5.619651252752485,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 23606
    },
    {
      "epoch": 5.6198893054811645,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.6607,
      "step": 23607
    },
    {
      "epoch": 5.620127358209843,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 23608
    },
    {
      "epoch": 5.620365410938523,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 23609
    },
    {
      "epoch": 5.620603463667202,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 23610
    },
    {
      "epoch": 5.620841516395882,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 23611
    },
    {
      "epoch": 5.621079569124561,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 23612
    },
    {
      "epoch": 5.62131762185324,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 23613
    },
    {
      "epoch": 5.62155567458192,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 23614
    },
    {
      "epoch": 5.6217937273106,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 23615
    },
    {
      "epoch": 5.622031780039279,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 23616
    },
    {
      "epoch": 5.622269832767958,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 23617
    },
    {
      "epoch": 5.6225078854966375,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 23618
    },
    {
      "epoch": 5.622745938225317,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 23619
    },
    {
      "epoch": 5.622983990953996,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 23620
    },
    {
      "epoch": 5.623222043682675,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 23621
    },
    {
      "epoch": 5.623460096411355,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 23622
    },
    {
      "epoch": 5.623698149140035,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 23623
    },
    {
      "epoch": 5.623936201868714,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 23624
    },
    {
      "epoch": 5.624174254597393,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 23625
    },
    {
      "epoch": 5.624412307326073,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 23626
    },
    {
      "epoch": 5.624650360054752,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 23627
    },
    {
      "epoch": 5.624888412783432,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3361,
      "step": 23628
    },
    {
      "epoch": 5.6251264655121105,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3995,
      "step": 23629
    },
    {
      "epoch": 5.62536451824079,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 23630
    },
    {
      "epoch": 5.62560257096947,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 23631
    },
    {
      "epoch": 5.625840623698149,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 23632
    },
    {
      "epoch": 5.626078676426829,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4001,
      "step": 23633
    },
    {
      "epoch": 5.626316729155508,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 23634
    },
    {
      "epoch": 5.626554781884187,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7635,
      "step": 23635
    },
    {
      "epoch": 5.626792834612867,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 23636
    },
    {
      "epoch": 5.627030887341546,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8674,
      "step": 23637
    },
    {
      "epoch": 5.627268940070225,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8403,
      "step": 23638
    },
    {
      "epoch": 5.627506992798905,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 23639
    },
    {
      "epoch": 5.627745045527584,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 23640
    },
    {
      "epoch": 5.627983098256264,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 23641
    },
    {
      "epoch": 5.628221150984944,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 23642
    },
    {
      "epoch": 5.628459203713622,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 23643
    },
    {
      "epoch": 5.628697256442302,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 23644
    },
    {
      "epoch": 5.6289353091709815,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 23645
    },
    {
      "epoch": 5.629173361899661,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 23646
    },
    {
      "epoch": 5.62941141462834,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 23647
    },
    {
      "epoch": 5.629649467357019,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 23648
    },
    {
      "epoch": 5.629887520085699,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 23649
    },
    {
      "epoch": 5.630125572814379,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2838,
      "step": 23650
    },
    {
      "epoch": 5.630363625543058,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 23651
    },
    {
      "epoch": 5.630601678271737,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 23652
    },
    {
      "epoch": 5.630839731000417,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 23653
    },
    {
      "epoch": 5.631077783729096,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 23654
    },
    {
      "epoch": 5.631315836457776,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3039,
      "step": 23655
    },
    {
      "epoch": 5.6315538891864545,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 23656
    },
    {
      "epoch": 5.631791941915134,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 23657
    },
    {
      "epoch": 5.632029994643814,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 23658
    },
    {
      "epoch": 5.632268047372493,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3424,
      "step": 23659
    },
    {
      "epoch": 5.632506100101172,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 23660
    },
    {
      "epoch": 5.632744152829852,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 23661
    },
    {
      "epoch": 5.632982205558531,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2716,
      "step": 23662
    },
    {
      "epoch": 5.633220258287211,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 23663
    },
    {
      "epoch": 5.63345831101589,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 23664
    },
    {
      "epoch": 5.633696363744569,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 23665
    },
    {
      "epoch": 5.633934416473249,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8378,
      "step": 23666
    },
    {
      "epoch": 5.634172469201928,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 23667
    },
    {
      "epoch": 5.634410521930608,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 23668
    },
    {
      "epoch": 5.634648574659287,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 23669
    },
    {
      "epoch": 5.634886627387966,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 23670
    },
    {
      "epoch": 5.635124680116646,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4687,
      "step": 23671
    },
    {
      "epoch": 5.6353627328453255,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 23672
    },
    {
      "epoch": 5.635600785574004,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 23673
    },
    {
      "epoch": 5.635838838302684,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 23674
    },
    {
      "epoch": 5.6360768910313634,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 23675
    },
    {
      "epoch": 5.636314943760043,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 23676
    },
    {
      "epoch": 5.636552996488723,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 23677
    },
    {
      "epoch": 5.636791049217401,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 23678
    },
    {
      "epoch": 5.637029101946081,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 23679
    },
    {
      "epoch": 5.637267154674761,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7523,
      "step": 23680
    },
    {
      "epoch": 5.63750520740344,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 23681
    },
    {
      "epoch": 5.637743260132119,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 23682
    },
    {
      "epoch": 5.6379813128607985,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 23683
    },
    {
      "epoch": 5.638219365589478,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 23684
    },
    {
      "epoch": 5.638457418318158,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8253,
      "step": 23685
    },
    {
      "epoch": 5.638695471046837,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 23686
    },
    {
      "epoch": 5.638933523775516,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 23687
    },
    {
      "epoch": 5.639171576504196,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 23688
    },
    {
      "epoch": 5.639409629232875,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 23689
    },
    {
      "epoch": 5.639647681961555,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 23690
    },
    {
      "epoch": 5.639885734690234,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 23691
    },
    {
      "epoch": 5.640123787418913,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9109,
      "step": 23692
    },
    {
      "epoch": 5.640361840147593,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 23693
    },
    {
      "epoch": 5.640599892876272,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 23694
    },
    {
      "epoch": 5.640837945604951,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 23695
    },
    {
      "epoch": 5.641075998333631,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.3324,
      "step": 23696
    },
    {
      "epoch": 5.64131405106231,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3597,
      "step": 23697
    },
    {
      "epoch": 5.64155210379099,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4774,
      "step": 23698
    },
    {
      "epoch": 5.641790156519669,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 23699
    },
    {
      "epoch": 5.642028209248348,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 23700
    },
    {
      "epoch": 5.642028209248348,
      "eval_runtime": 0.5254,
      "eval_samples_per_second": 1.903,
      "eval_steps_per_second": 1.903,
      "step": 23700
    },
    {
      "epoch": 5.642266261977028,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 23701
    },
    {
      "epoch": 5.6425043147057075,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 23702
    },
    {
      "epoch": 5.642742367434387,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2518,
      "step": 23703
    },
    {
      "epoch": 5.642980420163066,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 23704
    },
    {
      "epoch": 5.643218472891745,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 23705
    },
    {
      "epoch": 5.643456525620425,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 23706
    },
    {
      "epoch": 5.643694578349105,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 23707
    },
    {
      "epoch": 5.643932631077783,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 23708
    },
    {
      "epoch": 5.644170683806463,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3573,
      "step": 23709
    },
    {
      "epoch": 5.6444087365351425,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 23710
    },
    {
      "epoch": 5.644646789263822,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 23711
    },
    {
      "epoch": 5.644884841992502,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 23712
    },
    {
      "epoch": 5.6451228947211805,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 23713
    },
    {
      "epoch": 5.64536094744986,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 23714
    },
    {
      "epoch": 5.64559900017854,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 23715
    },
    {
      "epoch": 5.645837052907219,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 23716
    },
    {
      "epoch": 5.646075105635898,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 23717
    },
    {
      "epoch": 5.646313158364578,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 23718
    },
    {
      "epoch": 5.646551211093257,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 23719
    },
    {
      "epoch": 5.646789263821937,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0806,
      "step": 23720
    },
    {
      "epoch": 5.647027316550616,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 23721
    },
    {
      "epoch": 5.647265369279295,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 23722
    },
    {
      "epoch": 5.647503422007975,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 23723
    },
    {
      "epoch": 5.647741474736654,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 23724
    },
    {
      "epoch": 5.647979527465334,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 23725
    },
    {
      "epoch": 5.648217580194013,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3193,
      "step": 23726
    },
    {
      "epoch": 5.648455632922692,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 23727
    },
    {
      "epoch": 5.648693685651372,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 23728
    },
    {
      "epoch": 5.6489317383800515,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 23729
    },
    {
      "epoch": 5.649169791108731,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 23730
    },
    {
      "epoch": 5.64940784383741,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 23731
    },
    {
      "epoch": 5.649645896566089,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 23732
    },
    {
      "epoch": 5.649883949294769,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3968,
      "step": 23733
    },
    {
      "epoch": 5.650122002023448,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 23734
    },
    {
      "epoch": 5.650360054752127,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 23735
    },
    {
      "epoch": 5.650598107480807,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 23736
    },
    {
      "epoch": 5.6508361602094865,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 23737
    },
    {
      "epoch": 5.651074212938166,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 23738
    },
    {
      "epoch": 5.651312265666845,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 23739
    },
    {
      "epoch": 5.6515503183955245,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 23740
    },
    {
      "epoch": 5.651788371124204,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 23741
    },
    {
      "epoch": 5.652026423852884,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 23742
    },
    {
      "epoch": 5.652264476581562,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 23743
    },
    {
      "epoch": 5.652502529310242,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 23744
    },
    {
      "epoch": 5.652740582038922,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 23745
    },
    {
      "epoch": 5.652978634767601,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 23746
    },
    {
      "epoch": 5.653216687496281,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 23747
    },
    {
      "epoch": 5.6534547402249595,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 23748
    },
    {
      "epoch": 5.653692792953639,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3786,
      "step": 23749
    },
    {
      "epoch": 5.653930845682319,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 23750
    },
    {
      "epoch": 5.654168898410998,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8982,
      "step": 23751
    },
    {
      "epoch": 5.654406951139677,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 23752
    },
    {
      "epoch": 5.654645003868357,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 23753
    },
    {
      "epoch": 5.654883056597036,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 23754
    },
    {
      "epoch": 5.655121109325716,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 23755
    },
    {
      "epoch": 5.6553591620543955,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 23756
    },
    {
      "epoch": 5.655597214783074,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 23757
    },
    {
      "epoch": 5.655835267511754,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 23758
    },
    {
      "epoch": 5.656073320240433,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 23759
    },
    {
      "epoch": 5.656311372969113,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 23760
    },
    {
      "epoch": 5.656549425697792,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 23761
    },
    {
      "epoch": 5.656787478426471,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9432,
      "step": 23762
    },
    {
      "epoch": 5.657025531155151,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 23763
    },
    {
      "epoch": 5.6572635838838305,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 23764
    },
    {
      "epoch": 5.65750163661251,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 23765
    },
    {
      "epoch": 5.657739689341189,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 23766
    },
    {
      "epoch": 5.6579777420698685,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 23767
    },
    {
      "epoch": 5.658215794798548,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 23768
    },
    {
      "epoch": 5.658453847527227,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3789,
      "step": 23769
    },
    {
      "epoch": 5.658691900255906,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 23770
    },
    {
      "epoch": 5.658929952984586,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 23771
    },
    {
      "epoch": 5.659168005713266,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 23772
    },
    {
      "epoch": 5.659406058441945,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 23773
    },
    {
      "epoch": 5.659644111170624,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 23774
    },
    {
      "epoch": 5.6598821638993035,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 23775
    },
    {
      "epoch": 5.660120216627983,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 23776
    },
    {
      "epoch": 5.660358269356663,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 23777
    },
    {
      "epoch": 5.6605963220853415,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 23778
    },
    {
      "epoch": 5.660834374814021,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 23779
    },
    {
      "epoch": 5.661072427542701,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 23780
    },
    {
      "epoch": 5.66131048027138,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 23781
    },
    {
      "epoch": 5.66154853300006,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 23782
    },
    {
      "epoch": 5.661786585728739,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 23783
    },
    {
      "epoch": 5.662024638457418,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 23784
    },
    {
      "epoch": 5.662262691186098,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 23785
    },
    {
      "epoch": 5.662500743914777,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 23786
    },
    {
      "epoch": 5.662738796643456,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 23787
    },
    {
      "epoch": 5.662976849372136,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 23788
    },
    {
      "epoch": 5.663214902100815,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9832,
      "step": 23789
    },
    {
      "epoch": 5.663452954829495,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 23790
    },
    {
      "epoch": 5.663691007558175,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3768,
      "step": 23791
    },
    {
      "epoch": 5.663929060286853,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 23792
    },
    {
      "epoch": 5.664167113015533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 23793
    },
    {
      "epoch": 5.6644051657442125,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 23794
    },
    {
      "epoch": 5.664643218472892,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 23795
    },
    {
      "epoch": 5.664881271201571,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3698,
      "step": 23796
    },
    {
      "epoch": 5.66511932393025,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 23797
    },
    {
      "epoch": 5.66535737665893,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 23798
    },
    {
      "epoch": 5.66559542938761,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 23799
    },
    {
      "epoch": 5.665833482116289,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 23800
    },
    {
      "epoch": 5.665833482116289,
      "eval_runtime": 0.5055,
      "eval_samples_per_second": 1.978,
      "eval_steps_per_second": 1.978,
      "step": 23800
    },
    {
      "epoch": 5.666071534844968,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 23801
    },
    {
      "epoch": 5.6663095875736476,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 23802
    },
    {
      "epoch": 5.666547640302327,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1613,
      "step": 23803
    },
    {
      "epoch": 5.666785693031006,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 23804
    },
    {
      "epoch": 5.6670237457596855,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 23805
    },
    {
      "epoch": 5.667261798488365,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 23806
    },
    {
      "epoch": 5.667499851217045,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 23807
    },
    {
      "epoch": 5.667737903945724,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 23808
    },
    {
      "epoch": 5.667975956674403,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 23809
    },
    {
      "epoch": 5.668214009403083,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 23810
    },
    {
      "epoch": 5.668452062131762,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 23811
    },
    {
      "epoch": 5.668690114860442,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 23812
    },
    {
      "epoch": 5.6689281675891205,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 23813
    },
    {
      "epoch": 5.6691662203178,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 23814
    },
    {
      "epoch": 5.66940427304648,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0299,
      "step": 23815
    },
    {
      "epoch": 5.669642325775159,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 23816
    },
    {
      "epoch": 5.669880378503839,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 23817
    },
    {
      "epoch": 5.670118431232518,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 23818
    },
    {
      "epoch": 5.670356483961197,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 23819
    },
    {
      "epoch": 5.670594536689877,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 23820
    },
    {
      "epoch": 5.6708325894185565,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 23821
    },
    {
      "epoch": 5.671070642147235,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 23822
    },
    {
      "epoch": 5.671308694875915,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 23823
    },
    {
      "epoch": 5.671546747604594,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 23824
    },
    {
      "epoch": 5.671784800333274,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 23825
    },
    {
      "epoch": 5.672022853061954,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 23826
    },
    {
      "epoch": 5.672260905790632,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 23827
    },
    {
      "epoch": 5.672498958519312,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 23828
    },
    {
      "epoch": 5.672737011247992,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7761,
      "step": 23829
    },
    {
      "epoch": 5.672975063976671,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 23830
    },
    {
      "epoch": 5.67321311670535,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 23831
    },
    {
      "epoch": 5.6734511694340295,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 23832
    },
    {
      "epoch": 5.673689222162709,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4059,
      "step": 23833
    },
    {
      "epoch": 5.673927274891389,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 23834
    },
    {
      "epoch": 5.674165327620068,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 23835
    },
    {
      "epoch": 5.674403380348747,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 23836
    },
    {
      "epoch": 5.674641433077427,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 23837
    },
    {
      "epoch": 5.674879485806106,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 23838
    },
    {
      "epoch": 5.675117538534785,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 23839
    },
    {
      "epoch": 5.675355591263465,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 23840
    },
    {
      "epoch": 5.675593643992144,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 23841
    },
    {
      "epoch": 5.675831696720824,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 23842
    },
    {
      "epoch": 5.676069749449503,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 23843
    },
    {
      "epoch": 5.676307802178182,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 23844
    },
    {
      "epoch": 5.676545854906862,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 23845
    },
    {
      "epoch": 5.676783907635541,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 23846
    },
    {
      "epoch": 5.677021960364221,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 23847
    },
    {
      "epoch": 5.6772600130929,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 23848
    },
    {
      "epoch": 5.677498065821579,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 23849
    },
    {
      "epoch": 5.677736118550259,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 23850
    },
    {
      "epoch": 5.677974171278938,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 23851
    },
    {
      "epoch": 5.678212224007618,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3587,
      "step": 23852
    },
    {
      "epoch": 5.678450276736297,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 23853
    },
    {
      "epoch": 5.678688329464976,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 23854
    },
    {
      "epoch": 5.678926382193656,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 23855
    },
    {
      "epoch": 5.679164434922336,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 23856
    },
    {
      "epoch": 5.679402487651014,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 23857
    },
    {
      "epoch": 5.679640540379694,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 23858
    },
    {
      "epoch": 5.6798785931083735,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 23859
    },
    {
      "epoch": 5.680116645837053,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3588,
      "step": 23860
    },
    {
      "epoch": 5.680354698565733,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 23861
    },
    {
      "epoch": 5.680592751294411,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 23862
    },
    {
      "epoch": 5.680830804023091,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.5612,
      "step": 23863
    },
    {
      "epoch": 5.681068856751771,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 23864
    },
    {
      "epoch": 5.68130690948045,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3354,
      "step": 23865
    },
    {
      "epoch": 5.681544962209129,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 23866
    },
    {
      "epoch": 5.681783014937809,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 23867
    },
    {
      "epoch": 5.682021067666488,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 23868
    },
    {
      "epoch": 5.682259120395168,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 23869
    },
    {
      "epoch": 5.682497173123847,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 23870
    },
    {
      "epoch": 5.682735225852526,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 23871
    },
    {
      "epoch": 5.682973278581206,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 23872
    },
    {
      "epoch": 5.683211331309885,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 23873
    },
    {
      "epoch": 5.683449384038565,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 23874
    },
    {
      "epoch": 5.683687436767244,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 23875
    },
    {
      "epoch": 5.683925489495923,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 23876
    },
    {
      "epoch": 5.684163542224603,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 23877
    },
    {
      "epoch": 5.6844015949532825,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 23878
    },
    {
      "epoch": 5.684639647681961,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8097,
      "step": 23879
    },
    {
      "epoch": 5.684877700410641,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 23880
    },
    {
      "epoch": 5.68511575313932,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 23881
    },
    {
      "epoch": 5.685353805868,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 23882
    },
    {
      "epoch": 5.685591858596679,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 23883
    },
    {
      "epoch": 5.685829911325358,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.8432,
      "step": 23884
    },
    {
      "epoch": 5.686067964054038,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3383,
      "step": 23885
    },
    {
      "epoch": 5.6863060167827175,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 23886
    },
    {
      "epoch": 5.686544069511397,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.5939,
      "step": 23887
    },
    {
      "epoch": 5.686782122240076,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 23888
    },
    {
      "epoch": 5.6870201749687554,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3315,
      "step": 23889
    },
    {
      "epoch": 5.687258227697435,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 23890
    },
    {
      "epoch": 5.687496280426115,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3812,
      "step": 23891
    },
    {
      "epoch": 5.687734333154793,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 23892
    },
    {
      "epoch": 5.687972385883473,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 23893
    },
    {
      "epoch": 5.688210438612153,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 23894
    },
    {
      "epoch": 5.688448491340832,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 23895
    },
    {
      "epoch": 5.688686544069512,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 23896
    },
    {
      "epoch": 5.6889245967981905,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 23897
    },
    {
      "epoch": 5.68916264952687,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 23898
    },
    {
      "epoch": 5.68940070225555,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 23899
    },
    {
      "epoch": 5.689638754984229,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3342,
      "step": 23900
    },
    {
      "epoch": 5.689638754984229,
      "eval_runtime": 0.5147,
      "eval_samples_per_second": 1.943,
      "eval_steps_per_second": 1.943,
      "step": 23900
    },
    {
      "epoch": 5.689876807712908,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 23901
    },
    {
      "epoch": 5.690114860441588,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 23902
    },
    {
      "epoch": 5.690352913170267,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4508,
      "step": 23903
    },
    {
      "epoch": 5.690590965898947,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 23904
    },
    {
      "epoch": 5.6908290186276265,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 23905
    },
    {
      "epoch": 5.691067071356305,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2477,
      "step": 23906
    },
    {
      "epoch": 5.691305124084985,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 23907
    },
    {
      "epoch": 5.691543176813664,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 23908
    },
    {
      "epoch": 5.691781229542344,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 23909
    },
    {
      "epoch": 5.692019282271023,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 23910
    },
    {
      "epoch": 5.692257334999702,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 23911
    },
    {
      "epoch": 5.692495387728382,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 23912
    },
    {
      "epoch": 5.6927334404570615,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 23913
    },
    {
      "epoch": 5.692971493185741,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 23914
    },
    {
      "epoch": 5.69320954591442,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 23915
    },
    {
      "epoch": 5.6934475986430995,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 23916
    },
    {
      "epoch": 5.693685651371779,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 23917
    },
    {
      "epoch": 5.693923704100458,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 23918
    },
    {
      "epoch": 5.694161756829137,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 23919
    },
    {
      "epoch": 5.694399809557817,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 23920
    },
    {
      "epoch": 5.694637862286497,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 23921
    },
    {
      "epoch": 5.694875915015176,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 23922
    },
    {
      "epoch": 5.695113967743855,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 23923
    },
    {
      "epoch": 5.6953520204725345,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 23924
    },
    {
      "epoch": 5.695590073201214,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 23925
    },
    {
      "epoch": 5.695828125929894,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 23926
    },
    {
      "epoch": 5.6960661786585725,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 23927
    },
    {
      "epoch": 5.696304231387252,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 23928
    },
    {
      "epoch": 5.696542284115932,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 23929
    },
    {
      "epoch": 5.696780336844611,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 23930
    },
    {
      "epoch": 5.697018389573291,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.853,
      "step": 23931
    },
    {
      "epoch": 5.69725644230197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 23932
    },
    {
      "epoch": 5.697494495030649,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 23933
    },
    {
      "epoch": 5.697732547759329,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 23934
    },
    {
      "epoch": 5.697970600488008,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 23935
    },
    {
      "epoch": 5.698208653216687,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 23936
    },
    {
      "epoch": 5.698446705945367,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 23937
    },
    {
      "epoch": 5.698684758674046,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 23938
    },
    {
      "epoch": 5.698922811402726,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7555,
      "step": 23939
    },
    {
      "epoch": 5.6991608641314055,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3498,
      "step": 23940
    },
    {
      "epoch": 5.699398916860084,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 23941
    },
    {
      "epoch": 5.699636969588764,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 23942
    },
    {
      "epoch": 5.6998750223174435,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 23943
    },
    {
      "epoch": 5.700113075046123,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 23944
    },
    {
      "epoch": 5.700351127774802,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 23945
    },
    {
      "epoch": 5.700589180503481,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.408,
      "step": 23946
    },
    {
      "epoch": 5.700827233232161,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 23947
    },
    {
      "epoch": 5.701065285960841,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 23948
    },
    {
      "epoch": 5.70130333868952,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 23949
    },
    {
      "epoch": 5.701541391418199,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.6294,
      "step": 23950
    },
    {
      "epoch": 5.7017794441468785,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 23951
    },
    {
      "epoch": 5.702017496875558,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 23952
    },
    {
      "epoch": 5.702255549604237,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 23953
    },
    {
      "epoch": 5.7024936023329165,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5835,
      "step": 23954
    },
    {
      "epoch": 5.702731655061596,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 23955
    },
    {
      "epoch": 5.702969707790276,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 23956
    },
    {
      "epoch": 5.703207760518955,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 23957
    },
    {
      "epoch": 5.703445813247634,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 23958
    },
    {
      "epoch": 5.703683865976314,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 23959
    },
    {
      "epoch": 5.703921918704993,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 23960
    },
    {
      "epoch": 5.704159971433673,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 23961
    },
    {
      "epoch": 5.7043980241623515,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3306,
      "step": 23962
    },
    {
      "epoch": 5.704636076891031,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 23963
    },
    {
      "epoch": 5.704874129619711,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 23964
    },
    {
      "epoch": 5.70511218234839,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 23965
    },
    {
      "epoch": 5.70535023507707,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 23966
    },
    {
      "epoch": 5.705588287805749,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 23967
    },
    {
      "epoch": 5.705826340534428,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 23968
    },
    {
      "epoch": 5.706064393263108,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 23969
    },
    {
      "epoch": 5.7063024459917875,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 23970
    },
    {
      "epoch": 5.706540498720466,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 23971
    },
    {
      "epoch": 5.706778551449146,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 23972
    },
    {
      "epoch": 5.707016604177825,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.342,
      "step": 23973
    },
    {
      "epoch": 5.707254656906505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 23974
    },
    {
      "epoch": 5.707492709635185,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 23975
    },
    {
      "epoch": 5.707730762363863,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 23976
    },
    {
      "epoch": 5.707968815092543,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 23977
    },
    {
      "epoch": 5.7082068678212226,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7111,
      "step": 23978
    },
    {
      "epoch": 5.708444920549902,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 23979
    },
    {
      "epoch": 5.708682973278581,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 23980
    },
    {
      "epoch": 5.7089210260072605,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 23981
    },
    {
      "epoch": 5.70915907873594,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 23982
    },
    {
      "epoch": 5.70939713146462,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 23983
    },
    {
      "epoch": 5.709635184193299,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 23984
    },
    {
      "epoch": 5.709873236921978,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 23985
    },
    {
      "epoch": 5.710111289650658,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1682,
      "step": 23986
    },
    {
      "epoch": 5.710349342379337,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4264,
      "step": 23987
    },
    {
      "epoch": 5.710587395108016,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 23988
    },
    {
      "epoch": 5.7108254478366955,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 23989
    },
    {
      "epoch": 5.711063500565375,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3054,
      "step": 23990
    },
    {
      "epoch": 5.711301553294055,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 23991
    },
    {
      "epoch": 5.711539606022734,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 23992
    },
    {
      "epoch": 5.711777658751413,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 23993
    },
    {
      "epoch": 5.712015711480093,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 23994
    },
    {
      "epoch": 5.712253764208772,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 23995
    },
    {
      "epoch": 5.712491816937452,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 23996
    },
    {
      "epoch": 5.712729869666131,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 23997
    },
    {
      "epoch": 5.71296792239481,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 23998
    },
    {
      "epoch": 5.71320597512349,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 23999
    },
    {
      "epoch": 5.713444027852169,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 24000
    },
    {
      "epoch": 5.713444027852169,
      "eval_runtime": 0.526,
      "eval_samples_per_second": 1.901,
      "eval_steps_per_second": 1.901,
      "step": 24000
    },
    {
      "epoch": 5.713682080580849,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3265,
      "step": 24001
    },
    {
      "epoch": 5.713920133309528,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 24002
    },
    {
      "epoch": 5.714158186038207,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 24003
    },
    {
      "epoch": 5.714396238766887,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 24004
    },
    {
      "epoch": 5.714634291495567,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 24005
    },
    {
      "epoch": 5.714872344224245,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 24006
    },
    {
      "epoch": 5.715110396952925,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 24007
    },
    {
      "epoch": 5.7153484496816045,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 24008
    },
    {
      "epoch": 5.715586502410284,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 24009
    },
    {
      "epoch": 5.715824555138964,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 24010
    },
    {
      "epoch": 5.716062607867642,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 24011
    },
    {
      "epoch": 5.716300660596322,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 24012
    },
    {
      "epoch": 5.716538713325002,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 24013
    },
    {
      "epoch": 5.716776766053681,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 24014
    },
    {
      "epoch": 5.71701481878236,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 24015
    },
    {
      "epoch": 5.71725287151104,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 24016
    },
    {
      "epoch": 5.717490924239719,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 24017
    },
    {
      "epoch": 5.717728976968399,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 24018
    },
    {
      "epoch": 5.717967029697078,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 24019
    },
    {
      "epoch": 5.718205082425757,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3931,
      "step": 24020
    },
    {
      "epoch": 5.718443135154437,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8351,
      "step": 24021
    },
    {
      "epoch": 5.718681187883116,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 24022
    },
    {
      "epoch": 5.718919240611795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2958,
      "step": 24023
    },
    {
      "epoch": 5.719157293340475,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 24024
    },
    {
      "epoch": 5.719395346069154,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 24025
    },
    {
      "epoch": 5.719633398797834,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 24026
    },
    {
      "epoch": 5.719871451526513,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 24027
    },
    {
      "epoch": 5.720109504255192,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 24028
    },
    {
      "epoch": 5.720347556983872,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 24029
    },
    {
      "epoch": 5.720585609712551,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 24030
    },
    {
      "epoch": 5.720823662441231,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 24031
    },
    {
      "epoch": 5.72106171516991,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2115,
      "step": 24032
    },
    {
      "epoch": 5.721299767898589,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 24033
    },
    {
      "epoch": 5.721537820627269,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 24034
    },
    {
      "epoch": 5.7217758733559485,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 24035
    },
    {
      "epoch": 5.722013926084628,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 24036
    },
    {
      "epoch": 5.722251978813307,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 24037
    },
    {
      "epoch": 5.722490031541986,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 24038
    },
    {
      "epoch": 5.722728084270666,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 24039
    },
    {
      "epoch": 5.722966136999346,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9653,
      "step": 24040
    },
    {
      "epoch": 5.723204189728024,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 24041
    },
    {
      "epoch": 5.723442242456704,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 24042
    },
    {
      "epoch": 5.723680295185384,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 24043
    },
    {
      "epoch": 5.723918347914063,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 24044
    },
    {
      "epoch": 5.724156400642743,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 24045
    },
    {
      "epoch": 5.7243944533714215,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 24046
    },
    {
      "epoch": 5.724632506100101,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 24047
    },
    {
      "epoch": 5.724870558828781,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 24048
    },
    {
      "epoch": 5.72510861155746,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4604,
      "step": 24049
    },
    {
      "epoch": 5.725346664286139,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 24050
    },
    {
      "epoch": 5.725584717014819,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.423,
      "step": 24051
    },
    {
      "epoch": 5.725822769743498,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2992,
      "step": 24052
    },
    {
      "epoch": 5.726060822472178,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 24053
    },
    {
      "epoch": 5.7262988752008575,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 24054
    },
    {
      "epoch": 5.726536927929536,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 24055
    },
    {
      "epoch": 5.726774980658216,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 24056
    },
    {
      "epoch": 5.727013033386895,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 24057
    },
    {
      "epoch": 5.727251086115575,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 24058
    },
    {
      "epoch": 5.727489138844254,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 24059
    },
    {
      "epoch": 5.727727191572933,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 24060
    },
    {
      "epoch": 5.727965244301613,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4134,
      "step": 24061
    },
    {
      "epoch": 5.7282032970302925,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 24062
    },
    {
      "epoch": 5.728441349758971,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 24063
    },
    {
      "epoch": 5.728679402487651,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 24064
    },
    {
      "epoch": 5.7289174552163304,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 24065
    },
    {
      "epoch": 5.72915550794501,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 24066
    },
    {
      "epoch": 5.729393560673689,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 24067
    },
    {
      "epoch": 5.729631613402368,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 24068
    },
    {
      "epoch": 5.729869666131048,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 24069
    },
    {
      "epoch": 5.730107718859728,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 24070
    },
    {
      "epoch": 5.730345771588407,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 24071
    },
    {
      "epoch": 5.730583824317086,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 24072
    },
    {
      "epoch": 5.7308218770457655,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 24073
    },
    {
      "epoch": 5.731059929774445,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 24074
    },
    {
      "epoch": 5.731297982503125,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 24075
    },
    {
      "epoch": 5.731536035231803,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 24076
    },
    {
      "epoch": 5.731774087960483,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 24077
    },
    {
      "epoch": 5.732012140689163,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 24078
    },
    {
      "epoch": 5.732250193417842,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 24079
    },
    {
      "epoch": 5.732488246146522,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7136,
      "step": 24080
    },
    {
      "epoch": 5.732726298875201,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3124,
      "step": 24081
    },
    {
      "epoch": 5.73296435160388,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 24082
    },
    {
      "epoch": 5.73320240433256,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 24083
    },
    {
      "epoch": 5.733440457061239,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 24084
    },
    {
      "epoch": 5.733678509789918,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 24085
    },
    {
      "epoch": 5.733916562518598,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 24086
    },
    {
      "epoch": 5.734154615247277,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 24087
    },
    {
      "epoch": 5.734392667975957,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 24088
    },
    {
      "epoch": 5.7346307207046365,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 24089
    },
    {
      "epoch": 5.734868773433315,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 24090
    },
    {
      "epoch": 5.735106826161995,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3832,
      "step": 24091
    },
    {
      "epoch": 5.7353448788906745,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 24092
    },
    {
      "epoch": 5.735582931619354,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 24093
    },
    {
      "epoch": 5.735820984348033,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 24094
    },
    {
      "epoch": 5.736059037076712,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 24095
    },
    {
      "epoch": 5.736297089805392,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 24096
    },
    {
      "epoch": 5.736535142534072,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 24097
    },
    {
      "epoch": 5.73677319526275,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 24098
    },
    {
      "epoch": 5.73701124799143,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 24099
    },
    {
      "epoch": 5.7372493007201095,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 24100
    },
    {
      "epoch": 5.7372493007201095,
      "eval_runtime": 0.5329,
      "eval_samples_per_second": 1.877,
      "eval_steps_per_second": 1.877,
      "step": 24100
    },
    {
      "epoch": 5.737487353448789,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 24101
    },
    {
      "epoch": 5.737725406177468,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 24102
    },
    {
      "epoch": 5.7379634589061475,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 24103
    },
    {
      "epoch": 5.738201511634827,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 24104
    },
    {
      "epoch": 5.738439564363507,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 24105
    },
    {
      "epoch": 5.738677617092186,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 24106
    },
    {
      "epoch": 5.738915669820865,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 24107
    },
    {
      "epoch": 5.739153722549545,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 24108
    },
    {
      "epoch": 5.739391775278224,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 24109
    },
    {
      "epoch": 5.739629828006904,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 24110
    },
    {
      "epoch": 5.7398678807355825,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 24111
    },
    {
      "epoch": 5.740105933464262,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4221,
      "step": 24112
    },
    {
      "epoch": 5.740343986192942,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.358,
      "step": 24113
    },
    {
      "epoch": 5.740582038921621,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 24114
    },
    {
      "epoch": 5.740820091650301,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 24115
    },
    {
      "epoch": 5.74105814437898,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 24116
    },
    {
      "epoch": 5.741296197107659,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 24117
    },
    {
      "epoch": 5.741534249836339,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 24118
    },
    {
      "epoch": 5.7417723025650185,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 24119
    },
    {
      "epoch": 5.742010355293697,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 24120
    },
    {
      "epoch": 5.742248408022377,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 24121
    },
    {
      "epoch": 5.742486460751056,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 24122
    },
    {
      "epoch": 5.742724513479736,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 24123
    },
    {
      "epoch": 5.742962566208416,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 24124
    },
    {
      "epoch": 5.743200618937094,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 24125
    },
    {
      "epoch": 5.743438671665774,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1647,
      "step": 24126
    },
    {
      "epoch": 5.7436767243944535,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 24127
    },
    {
      "epoch": 5.743914777123133,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 24128
    },
    {
      "epoch": 5.744152829851812,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 24129
    },
    {
      "epoch": 5.7443908825804915,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 24130
    },
    {
      "epoch": 5.744628935309171,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 24131
    },
    {
      "epoch": 5.744866988037851,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 24132
    },
    {
      "epoch": 5.74510504076653,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 24133
    },
    {
      "epoch": 5.745343093495209,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 24134
    },
    {
      "epoch": 5.745581146223889,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 24135
    },
    {
      "epoch": 5.745819198952568,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 24136
    },
    {
      "epoch": 5.746057251681247,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 24137
    },
    {
      "epoch": 5.7462953044099265,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 24138
    },
    {
      "epoch": 5.746533357138606,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3907,
      "step": 24139
    },
    {
      "epoch": 5.746771409867286,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 24140
    },
    {
      "epoch": 5.747009462595965,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 24141
    },
    {
      "epoch": 5.747247515324644,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 24142
    },
    {
      "epoch": 5.747485568053324,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 24143
    },
    {
      "epoch": 5.747723620782003,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 24144
    },
    {
      "epoch": 5.747961673510683,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 24145
    },
    {
      "epoch": 5.748199726239362,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 24146
    },
    {
      "epoch": 5.748437778968041,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 24147
    },
    {
      "epoch": 5.748675831696721,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 24148
    },
    {
      "epoch": 5.7489138844254,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8197,
      "step": 24149
    },
    {
      "epoch": 5.74915193715408,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 24150
    },
    {
      "epoch": 5.749389989882759,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 24151
    },
    {
      "epoch": 5.749628042611438,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1923,
      "step": 24152
    },
    {
      "epoch": 5.749866095340118,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 24153
    },
    {
      "epoch": 5.7501041480687975,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 24154
    },
    {
      "epoch": 5.750342200797476,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 24155
    },
    {
      "epoch": 5.750580253526156,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 24156
    },
    {
      "epoch": 5.7508183062548355,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 24157
    },
    {
      "epoch": 5.751056358983515,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 24158
    },
    {
      "epoch": 5.751294411712195,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 24159
    },
    {
      "epoch": 5.751532464440873,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 24160
    },
    {
      "epoch": 5.751770517169553,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3998,
      "step": 24161
    },
    {
      "epoch": 5.752008569898233,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 24162
    },
    {
      "epoch": 5.752246622626912,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 24163
    },
    {
      "epoch": 5.752484675355591,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 24164
    },
    {
      "epoch": 5.7527227280842705,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 24165
    },
    {
      "epoch": 5.75296078081295,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 24166
    },
    {
      "epoch": 5.75319883354163,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 24167
    },
    {
      "epoch": 5.753436886270309,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 24168
    },
    {
      "epoch": 5.753674938998988,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 24169
    },
    {
      "epoch": 5.753912991727668,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 24170
    },
    {
      "epoch": 5.754151044456347,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 24171
    },
    {
      "epoch": 5.754389097185026,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 24172
    },
    {
      "epoch": 5.754627149913706,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4493,
      "step": 24173
    },
    {
      "epoch": 5.754865202642385,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 24174
    },
    {
      "epoch": 5.755103255371065,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.4249,
      "step": 24175
    },
    {
      "epoch": 5.755341308099744,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 24176
    },
    {
      "epoch": 5.755579360828423,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 24177
    },
    {
      "epoch": 5.755817413557103,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 24178
    },
    {
      "epoch": 5.756055466285782,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 24179
    },
    {
      "epoch": 5.756293519014462,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 24180
    },
    {
      "epoch": 5.756531571743141,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 24181
    },
    {
      "epoch": 5.75676962447182,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 24182
    },
    {
      "epoch": 5.7570076772005,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 24183
    },
    {
      "epoch": 5.7572457299291795,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 24184
    },
    {
      "epoch": 5.757483782657859,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 24185
    },
    {
      "epoch": 5.757721835386538,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2886,
      "step": 24186
    },
    {
      "epoch": 5.757959888115217,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 24187
    },
    {
      "epoch": 5.758197940843897,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 24188
    },
    {
      "epoch": 5.758435993572577,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 24189
    },
    {
      "epoch": 5.758674046301255,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 24190
    },
    {
      "epoch": 5.758912099029935,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 24191
    },
    {
      "epoch": 5.7591501517586146,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 24192
    },
    {
      "epoch": 5.759388204487294,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 24193
    },
    {
      "epoch": 5.759626257215974,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 24194
    },
    {
      "epoch": 5.7598643099446525,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 24195
    },
    {
      "epoch": 5.760102362673332,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3648,
      "step": 24196
    },
    {
      "epoch": 5.760340415402012,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 24197
    },
    {
      "epoch": 5.760578468130691,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 24198
    },
    {
      "epoch": 5.76081652085937,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 24199
    },
    {
      "epoch": 5.76105457358805,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 24200
    },
    {
      "epoch": 5.76105457358805,
      "eval_runtime": 0.5281,
      "eval_samples_per_second": 1.893,
      "eval_steps_per_second": 1.893,
      "step": 24200
    },
    {
      "epoch": 5.761292626316729,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 24201
    },
    {
      "epoch": 5.761530679045409,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 24202
    },
    {
      "epoch": 5.761768731774088,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 24203
    },
    {
      "epoch": 5.762006784502767,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 24204
    },
    {
      "epoch": 5.762244837231447,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2772,
      "step": 24205
    },
    {
      "epoch": 5.762482889960126,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 24206
    },
    {
      "epoch": 5.762720942688805,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 24207
    },
    {
      "epoch": 5.762958995417485,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 24208
    },
    {
      "epoch": 5.763197048146164,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 24209
    },
    {
      "epoch": 5.763435100874844,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 24210
    },
    {
      "epoch": 5.7636731536035235,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 24211
    },
    {
      "epoch": 5.763911206332202,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 24212
    },
    {
      "epoch": 5.764149259060882,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 24213
    },
    {
      "epoch": 5.764387311789561,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 24214
    },
    {
      "epoch": 5.764625364518241,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8629,
      "step": 24215
    },
    {
      "epoch": 5.76486341724692,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 24216
    },
    {
      "epoch": 5.765101469975599,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 24217
    },
    {
      "epoch": 5.765339522704279,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.3272,
      "step": 24218
    },
    {
      "epoch": 5.765577575432959,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 24219
    },
    {
      "epoch": 5.765815628161638,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3452,
      "step": 24220
    },
    {
      "epoch": 5.766053680890317,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 24221
    },
    {
      "epoch": 5.7662917336189965,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 24222
    },
    {
      "epoch": 5.766529786347676,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 24223
    },
    {
      "epoch": 5.766767839076356,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 24224
    },
    {
      "epoch": 5.767005891805034,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 24225
    },
    {
      "epoch": 5.767243944533714,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 24226
    },
    {
      "epoch": 5.767481997262394,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 24227
    },
    {
      "epoch": 5.767720049991073,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 24228
    },
    {
      "epoch": 5.767958102719753,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 24229
    },
    {
      "epoch": 5.768196155448432,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 24230
    },
    {
      "epoch": 5.768434208177111,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 24231
    },
    {
      "epoch": 5.768672260905791,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 24232
    },
    {
      "epoch": 5.76891031363447,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 24233
    },
    {
      "epoch": 5.769148366363149,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 24234
    },
    {
      "epoch": 5.769386419091829,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 24235
    },
    {
      "epoch": 5.769624471820508,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 24236
    },
    {
      "epoch": 5.769862524549188,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 24237
    },
    {
      "epoch": 5.7701005772778675,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 24238
    },
    {
      "epoch": 5.770338630006546,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 24239
    },
    {
      "epoch": 5.770576682735226,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 24240
    },
    {
      "epoch": 5.770814735463905,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 24241
    },
    {
      "epoch": 5.771052788192584,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 24242
    },
    {
      "epoch": 5.771290840921264,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 24243
    },
    {
      "epoch": 5.771528893649943,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7857,
      "step": 24244
    },
    {
      "epoch": 5.771766946378623,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 24245
    },
    {
      "epoch": 5.772004999107303,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 24246
    },
    {
      "epoch": 5.772243051835981,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 24247
    },
    {
      "epoch": 5.772481104564661,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 24248
    },
    {
      "epoch": 5.7727191572933405,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 24249
    },
    {
      "epoch": 5.77295721002202,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 24250
    },
    {
      "epoch": 5.773195262750699,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 24251
    },
    {
      "epoch": 5.773433315479378,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 24252
    },
    {
      "epoch": 5.773671368208058,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 24253
    },
    {
      "epoch": 5.773909420936738,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 24254
    },
    {
      "epoch": 5.774147473665417,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 24255
    },
    {
      "epoch": 5.774385526394096,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 24256
    },
    {
      "epoch": 5.774623579122776,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 24257
    },
    {
      "epoch": 5.774861631851455,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 24258
    },
    {
      "epoch": 5.775099684580135,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 24259
    },
    {
      "epoch": 5.7753377373088135,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2793,
      "step": 24260
    },
    {
      "epoch": 5.775575790037493,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 24261
    },
    {
      "epoch": 5.775813842766173,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 24262
    },
    {
      "epoch": 5.776051895494852,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 24263
    },
    {
      "epoch": 5.776289948223532,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4229,
      "step": 24264
    },
    {
      "epoch": 5.776528000952211,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8957,
      "step": 24265
    },
    {
      "epoch": 5.77676605368089,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 24266
    },
    {
      "epoch": 5.77700410640957,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 24267
    },
    {
      "epoch": 5.7772421591382495,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3562,
      "step": 24268
    },
    {
      "epoch": 5.777480211866928,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 24269
    },
    {
      "epoch": 5.777718264595608,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 24270
    },
    {
      "epoch": 5.777956317324287,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3183,
      "step": 24271
    },
    {
      "epoch": 5.778194370052967,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3256,
      "step": 24272
    },
    {
      "epoch": 5.778432422781647,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 24273
    },
    {
      "epoch": 5.778670475510325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 24274
    },
    {
      "epoch": 5.778908528239005,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 24275
    },
    {
      "epoch": 5.7791465809676845,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 24276
    },
    {
      "epoch": 5.779384633696364,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 24277
    },
    {
      "epoch": 5.779622686425043,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 24278
    },
    {
      "epoch": 5.7798607391537224,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1576,
      "step": 24279
    },
    {
      "epoch": 5.780098791882402,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 24280
    },
    {
      "epoch": 5.780336844611082,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 24281
    },
    {
      "epoch": 5.78057489733976,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 24282
    },
    {
      "epoch": 5.78081295006844,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 24283
    },
    {
      "epoch": 5.78105100279712,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 24284
    },
    {
      "epoch": 5.781289055525799,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 24285
    },
    {
      "epoch": 5.781527108254478,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7006,
      "step": 24286
    },
    {
      "epoch": 5.7817651609831575,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3741,
      "step": 24287
    },
    {
      "epoch": 5.782003213711837,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 24288
    },
    {
      "epoch": 5.782241266440517,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 24289
    },
    {
      "epoch": 5.782479319169196,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 24290
    },
    {
      "epoch": 5.782717371897875,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 24291
    },
    {
      "epoch": 5.782955424626555,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 24292
    },
    {
      "epoch": 5.783193477355234,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 24293
    },
    {
      "epoch": 5.783431530083914,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 24294
    },
    {
      "epoch": 5.783669582812593,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 24295
    },
    {
      "epoch": 5.783907635541272,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 24296
    },
    {
      "epoch": 5.784145688269952,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 24297
    },
    {
      "epoch": 5.784383740998631,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 24298
    },
    {
      "epoch": 5.784621793727311,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 24299
    },
    {
      "epoch": 5.78485984645599,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 24300
    },
    {
      "epoch": 5.78485984645599,
      "eval_runtime": 0.5411,
      "eval_samples_per_second": 1.848,
      "eval_steps_per_second": 1.848,
      "step": 24300
    },
    {
      "epoch": 5.785097899184669,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 24301
    },
    {
      "epoch": 5.785335951913349,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 24302
    },
    {
      "epoch": 5.7855740046420285,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 24303
    },
    {
      "epoch": 5.785812057370707,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 24304
    },
    {
      "epoch": 5.786050110099387,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 24305
    },
    {
      "epoch": 5.7862881628280665,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 24306
    },
    {
      "epoch": 5.786526215556746,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 24307
    },
    {
      "epoch": 5.786764268285426,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 24308
    },
    {
      "epoch": 5.787002321014104,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 24309
    },
    {
      "epoch": 5.787240373742784,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.6038,
      "step": 24310
    },
    {
      "epoch": 5.787478426471464,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 24311
    },
    {
      "epoch": 5.787716479200143,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 24312
    },
    {
      "epoch": 5.787954531928822,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 24313
    },
    {
      "epoch": 5.7881925846575015,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 24314
    },
    {
      "epoch": 5.788430637386181,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 24315
    },
    {
      "epoch": 5.788668690114861,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 24316
    },
    {
      "epoch": 5.78890674284354,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 24317
    },
    {
      "epoch": 5.789144795572219,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3457,
      "step": 24318
    },
    {
      "epoch": 5.789382848300899,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 24319
    },
    {
      "epoch": 5.789620901029578,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 24320
    },
    {
      "epoch": 5.789858953758257,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9487,
      "step": 24321
    },
    {
      "epoch": 5.790097006486937,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 24322
    },
    {
      "epoch": 5.790335059215616,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 24323
    },
    {
      "epoch": 5.790573111944296,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 24324
    },
    {
      "epoch": 5.790811164672975,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 24325
    },
    {
      "epoch": 5.791049217401654,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 24326
    },
    {
      "epoch": 5.791287270130334,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 24327
    },
    {
      "epoch": 5.791525322859013,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 24328
    },
    {
      "epoch": 5.791763375587693,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 24329
    },
    {
      "epoch": 5.792001428316372,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2836,
      "step": 24330
    },
    {
      "epoch": 5.792239481045051,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 24331
    },
    {
      "epoch": 5.792477533773731,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 24332
    },
    {
      "epoch": 5.7927155865024105,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 24333
    },
    {
      "epoch": 5.79295363923109,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 24334
    },
    {
      "epoch": 5.793191691959769,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 24335
    },
    {
      "epoch": 5.793429744688448,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 24336
    },
    {
      "epoch": 5.793667797417128,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3772,
      "step": 24337
    },
    {
      "epoch": 5.793905850145808,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 24338
    },
    {
      "epoch": 5.794143902874486,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 24339
    },
    {
      "epoch": 5.794381955603166,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.6953,
      "step": 24340
    },
    {
      "epoch": 5.7946200083318455,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 24341
    },
    {
      "epoch": 5.794858061060525,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3045,
      "step": 24342
    },
    {
      "epoch": 5.795096113789205,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 24343
    },
    {
      "epoch": 5.7953341665178835,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 24344
    },
    {
      "epoch": 5.795572219246563,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 24345
    },
    {
      "epoch": 5.795810271975243,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 24346
    },
    {
      "epoch": 5.796048324703922,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 24347
    },
    {
      "epoch": 5.796286377432601,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 24348
    },
    {
      "epoch": 5.796524430161281,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 24349
    },
    {
      "epoch": 5.79676248288996,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 24350
    },
    {
      "epoch": 5.79700053561864,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 24351
    },
    {
      "epoch": 5.797238588347319,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 24352
    },
    {
      "epoch": 5.797476641075998,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 24353
    },
    {
      "epoch": 5.797714693804678,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 24354
    },
    {
      "epoch": 5.797952746533357,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 24355
    },
    {
      "epoch": 5.798190799262036,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 24356
    },
    {
      "epoch": 5.798428851990716,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 24357
    },
    {
      "epoch": 5.798666904719395,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 24358
    },
    {
      "epoch": 5.798904957448075,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 24359
    },
    {
      "epoch": 5.7991430101767545,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 24360
    },
    {
      "epoch": 5.799381062905433,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 24361
    },
    {
      "epoch": 5.799619115634113,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.4466,
      "step": 24362
    },
    {
      "epoch": 5.799857168362792,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 24363
    },
    {
      "epoch": 5.800095221091472,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2678,
      "step": 24364
    },
    {
      "epoch": 5.800333273820151,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 24365
    },
    {
      "epoch": 5.80057132654883,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 24366
    },
    {
      "epoch": 5.80080937927751,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 24367
    },
    {
      "epoch": 5.8010474320061896,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 24368
    },
    {
      "epoch": 5.801285484734869,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 24369
    },
    {
      "epoch": 5.801523537463548,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 24370
    },
    {
      "epoch": 5.8017615901922275,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 24371
    },
    {
      "epoch": 5.801999642920907,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 24372
    },
    {
      "epoch": 5.802237695649587,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 24373
    },
    {
      "epoch": 5.802475748378265,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 24374
    },
    {
      "epoch": 5.802713801106945,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 24375
    },
    {
      "epoch": 5.802951853835625,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 24376
    },
    {
      "epoch": 5.803189906564304,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4482,
      "step": 24377
    },
    {
      "epoch": 5.803427959292984,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 24378
    },
    {
      "epoch": 5.8036660120216625,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 24379
    },
    {
      "epoch": 5.803904064750342,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 24380
    },
    {
      "epoch": 5.804142117479022,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 24381
    },
    {
      "epoch": 5.804380170207701,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2962,
      "step": 24382
    },
    {
      "epoch": 5.80461822293638,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 24383
    },
    {
      "epoch": 5.80485627566506,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 24384
    },
    {
      "epoch": 5.805094328393739,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 24385
    },
    {
      "epoch": 5.805332381122419,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3039,
      "step": 24386
    },
    {
      "epoch": 5.8055704338510985,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3418,
      "step": 24387
    },
    {
      "epoch": 5.805808486579777,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3179,
      "step": 24388
    },
    {
      "epoch": 5.806046539308457,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1952,
      "step": 24389
    },
    {
      "epoch": 5.806284592037136,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 24390
    },
    {
      "epoch": 5.806522644765815,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 24391
    },
    {
      "epoch": 5.806760697494495,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2626,
      "step": 24392
    },
    {
      "epoch": 5.806998750223174,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 24393
    },
    {
      "epoch": 5.807236802951854,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 24394
    },
    {
      "epoch": 5.807474855680534,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 24395
    },
    {
      "epoch": 5.807712908409212,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 24396
    },
    {
      "epoch": 5.807950961137892,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.4682,
      "step": 24397
    },
    {
      "epoch": 5.8081890138665715,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 24398
    },
    {
      "epoch": 5.808427066595251,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 24399
    },
    {
      "epoch": 5.80866511932393,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.3171,
      "step": 24400
    },
    {
      "epoch": 5.80866511932393,
      "eval_runtime": 0.529,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 24400
    },
    {
      "epoch": 5.808903172052609,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 24401
    },
    {
      "epoch": 5.809141224781289,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 24402
    },
    {
      "epoch": 5.809379277509969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 24403
    },
    {
      "epoch": 5.809617330238648,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 24404
    },
    {
      "epoch": 5.809855382967327,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 24405
    },
    {
      "epoch": 5.810093435696007,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 24406
    },
    {
      "epoch": 5.810331488424686,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2993,
      "step": 24407
    },
    {
      "epoch": 5.810569541153366,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 24408
    },
    {
      "epoch": 5.8108075938820445,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 24409
    },
    {
      "epoch": 5.811045646610724,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 24410
    },
    {
      "epoch": 5.811283699339404,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 24411
    },
    {
      "epoch": 5.811521752068083,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9645,
      "step": 24412
    },
    {
      "epoch": 5.811759804796763,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 24413
    },
    {
      "epoch": 5.811997857525442,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 24414
    },
    {
      "epoch": 5.812235910254121,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3869,
      "step": 24415
    },
    {
      "epoch": 5.812473962982801,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2858,
      "step": 24416
    },
    {
      "epoch": 5.81271201571148,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 24417
    },
    {
      "epoch": 5.812950068440159,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 24418
    },
    {
      "epoch": 5.813188121168839,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 24419
    },
    {
      "epoch": 5.813426173897518,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 24420
    },
    {
      "epoch": 5.813664226626198,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 24421
    },
    {
      "epoch": 5.813902279354878,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1775,
      "step": 24422
    },
    {
      "epoch": 5.814140332083556,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 24423
    },
    {
      "epoch": 5.814378384812236,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 24424
    },
    {
      "epoch": 5.8146164375409155,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 24425
    },
    {
      "epoch": 5.814854490269594,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 24426
    },
    {
      "epoch": 5.815092542998274,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3122,
      "step": 24427
    },
    {
      "epoch": 5.815330595726953,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 24428
    },
    {
      "epoch": 5.815568648455633,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3679,
      "step": 24429
    },
    {
      "epoch": 5.815806701184313,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.191,
      "step": 24430
    },
    {
      "epoch": 5.816044753912991,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 24431
    },
    {
      "epoch": 5.816282806641671,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 24432
    },
    {
      "epoch": 5.816520859370351,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 24433
    },
    {
      "epoch": 5.81675891209903,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 24434
    },
    {
      "epoch": 5.816996964827709,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 24435
    },
    {
      "epoch": 5.8172350175563885,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 24436
    },
    {
      "epoch": 5.817473070285068,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 24437
    },
    {
      "epoch": 5.817711123013748,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 24438
    },
    {
      "epoch": 5.817949175742427,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 24439
    },
    {
      "epoch": 5.818187228471106,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 24440
    },
    {
      "epoch": 5.818425281199786,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2074,
      "step": 24441
    },
    {
      "epoch": 5.818663333928465,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3179,
      "step": 24442
    },
    {
      "epoch": 5.818901386657145,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 24443
    },
    {
      "epoch": 5.819139439385824,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 24444
    },
    {
      "epoch": 5.819377492114503,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9147,
      "step": 24445
    },
    {
      "epoch": 5.819615544843183,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 24446
    },
    {
      "epoch": 5.819853597571862,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3301,
      "step": 24447
    },
    {
      "epoch": 5.820091650300542,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 24448
    },
    {
      "epoch": 5.820329703029221,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2821,
      "step": 24449
    },
    {
      "epoch": 5.8205677557579,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 24450
    },
    {
      "epoch": 5.82080580848658,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 24451
    },
    {
      "epoch": 5.8210438612152595,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 24452
    },
    {
      "epoch": 5.821281913943938,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 24453
    },
    {
      "epoch": 5.821519966672618,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2913,
      "step": 24454
    },
    {
      "epoch": 5.8217580194012974,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 24455
    },
    {
      "epoch": 5.821996072129977,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 24456
    },
    {
      "epoch": 5.822234124858657,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 24457
    },
    {
      "epoch": 5.822472177587335,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 24458
    },
    {
      "epoch": 5.822710230316015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 24459
    },
    {
      "epoch": 5.822948283044695,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 24460
    },
    {
      "epoch": 5.823186335773374,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.598,
      "step": 24461
    },
    {
      "epoch": 5.823424388502053,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 24462
    },
    {
      "epoch": 5.8236624412307325,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 24463
    },
    {
      "epoch": 5.823900493959412,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 24464
    },
    {
      "epoch": 5.824138546688092,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 24465
    },
    {
      "epoch": 5.82437659941677,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 24466
    },
    {
      "epoch": 5.82461465214545,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 24467
    },
    {
      "epoch": 5.82485270487413,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 24468
    },
    {
      "epoch": 5.825090757602809,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 24469
    },
    {
      "epoch": 5.825328810331488,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 24470
    },
    {
      "epoch": 5.825566863060168,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 24471
    },
    {
      "epoch": 5.825804915788847,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 24472
    },
    {
      "epoch": 5.826042968517527,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 24473
    },
    {
      "epoch": 5.826281021246206,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 24474
    },
    {
      "epoch": 5.826519073974885,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 24475
    },
    {
      "epoch": 5.826757126703565,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 24476
    },
    {
      "epoch": 5.826995179432244,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 24477
    },
    {
      "epoch": 5.827233232160924,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 24478
    },
    {
      "epoch": 5.827471284889603,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 24479
    },
    {
      "epoch": 5.827709337618282,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 24480
    },
    {
      "epoch": 5.827947390346962,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 24481
    },
    {
      "epoch": 5.8281854430756415,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2466,
      "step": 24482
    },
    {
      "epoch": 5.828423495804321,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 24483
    },
    {
      "epoch": 5.828661548533,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 24484
    },
    {
      "epoch": 5.828899601261679,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 24485
    },
    {
      "epoch": 5.829137653990359,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 24486
    },
    {
      "epoch": 5.829375706719039,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 24487
    },
    {
      "epoch": 5.829613759447717,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 24488
    },
    {
      "epoch": 5.829851812176397,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 24489
    },
    {
      "epoch": 5.8300898649050765,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 24490
    },
    {
      "epoch": 5.830327917633756,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 24491
    },
    {
      "epoch": 5.830565970362436,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 24492
    },
    {
      "epoch": 5.8308040230911145,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3239,
      "step": 24493
    },
    {
      "epoch": 5.831042075819794,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 24494
    },
    {
      "epoch": 5.831280128548474,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 24495
    },
    {
      "epoch": 5.831518181277153,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 24496
    },
    {
      "epoch": 5.831756234005832,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3114,
      "step": 24497
    },
    {
      "epoch": 5.831994286734512,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 24498
    },
    {
      "epoch": 5.832232339463191,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0965,
      "step": 24499
    },
    {
      "epoch": 5.832470392191871,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 24500
    },
    {
      "epoch": 5.832470392191871,
      "eval_runtime": 0.5175,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 24500
    },
    {
      "epoch": 5.8327084449205495,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 24501
    },
    {
      "epoch": 5.832946497649229,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 24502
    },
    {
      "epoch": 5.833184550377909,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 24503
    },
    {
      "epoch": 5.833422603106588,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 24504
    },
    {
      "epoch": 5.833660655835267,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 24505
    },
    {
      "epoch": 5.833898708563947,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 24506
    },
    {
      "epoch": 5.834136761292626,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 24507
    },
    {
      "epoch": 5.834374814021306,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 24508
    },
    {
      "epoch": 5.8346128667499855,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 24509
    },
    {
      "epoch": 5.834850919478664,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 24510
    },
    {
      "epoch": 5.835088972207344,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2896,
      "step": 24511
    },
    {
      "epoch": 5.835327024936023,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 24512
    },
    {
      "epoch": 5.835565077664703,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 24513
    },
    {
      "epoch": 5.835803130393382,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 24514
    },
    {
      "epoch": 5.836041183122061,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 24515
    },
    {
      "epoch": 5.836279235850741,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 24516
    },
    {
      "epoch": 5.8365172885794205,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 24517
    },
    {
      "epoch": 5.8367553413081,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 24518
    },
    {
      "epoch": 5.836993394036779,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 24519
    },
    {
      "epoch": 5.8372314467654585,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 24520
    },
    {
      "epoch": 5.837469499494138,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 24521
    },
    {
      "epoch": 5.837707552222818,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 24522
    },
    {
      "epoch": 5.837945604951496,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 24523
    },
    {
      "epoch": 5.838183657680176,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 24524
    },
    {
      "epoch": 5.838421710408856,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 24525
    },
    {
      "epoch": 5.838659763137535,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 24526
    },
    {
      "epoch": 5.838897815866215,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 24527
    },
    {
      "epoch": 5.8391358685948935,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 24528
    },
    {
      "epoch": 5.839373921323573,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 24529
    },
    {
      "epoch": 5.839611974052253,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 24530
    },
    {
      "epoch": 5.839850026780932,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 24531
    },
    {
      "epoch": 5.840088079509611,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 24532
    },
    {
      "epoch": 5.840326132238291,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 24533
    },
    {
      "epoch": 5.84056418496697,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 24534
    },
    {
      "epoch": 5.84080223769565,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 24535
    },
    {
      "epoch": 5.8410402904243295,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 24536
    },
    {
      "epoch": 5.841278343153008,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 24537
    },
    {
      "epoch": 5.841516395881688,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 24538
    },
    {
      "epoch": 5.841754448610367,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3391,
      "step": 24539
    },
    {
      "epoch": 5.841992501339046,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 24540
    },
    {
      "epoch": 5.842230554067726,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 24541
    },
    {
      "epoch": 5.842468606796405,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4204,
      "step": 24542
    },
    {
      "epoch": 5.842706659525085,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 24543
    },
    {
      "epoch": 5.8429447122537645,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 24544
    },
    {
      "epoch": 5.843182764982443,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 24545
    },
    {
      "epoch": 5.843420817711123,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 24546
    },
    {
      "epoch": 5.8436588704398025,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 24547
    },
    {
      "epoch": 5.843896923168482,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.5027,
      "step": 24548
    },
    {
      "epoch": 5.844134975897161,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3057,
      "step": 24549
    },
    {
      "epoch": 5.84437302862584,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 24550
    },
    {
      "epoch": 5.84461108135452,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 24551
    },
    {
      "epoch": 5.8448491340832,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 24552
    },
    {
      "epoch": 5.845087186811879,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 24553
    },
    {
      "epoch": 5.845325239540558,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 24554
    },
    {
      "epoch": 5.8455632922692375,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0622,
      "step": 24555
    },
    {
      "epoch": 5.845801344997917,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 24556
    },
    {
      "epoch": 5.846039397726597,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 24557
    },
    {
      "epoch": 5.8462774504552755,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 24558
    },
    {
      "epoch": 5.846515503183955,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 24559
    },
    {
      "epoch": 5.846753555912635,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 24560
    },
    {
      "epoch": 5.846991608641314,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.89,
      "step": 24561
    },
    {
      "epoch": 5.847229661369994,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 24562
    },
    {
      "epoch": 5.847467714098673,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 24563
    },
    {
      "epoch": 5.847705766827352,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 24564
    },
    {
      "epoch": 5.847943819556032,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 24565
    },
    {
      "epoch": 5.848181872284711,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 24566
    },
    {
      "epoch": 5.84841992501339,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 24567
    },
    {
      "epoch": 5.84865797774207,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3617,
      "step": 24568
    },
    {
      "epoch": 5.848896030470749,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9399,
      "step": 24569
    },
    {
      "epoch": 5.849134083199429,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3237,
      "step": 24570
    },
    {
      "epoch": 5.849372135928109,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 24571
    },
    {
      "epoch": 5.849610188656787,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3159,
      "step": 24572
    },
    {
      "epoch": 5.849848241385467,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 24573
    },
    {
      "epoch": 5.8500862941141465,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 24574
    },
    {
      "epoch": 5.850324346842825,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 24575
    },
    {
      "epoch": 5.850562399571505,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 24576
    },
    {
      "epoch": 5.850800452300184,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 24577
    },
    {
      "epoch": 5.851038505028864,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 24578
    },
    {
      "epoch": 5.851276557757544,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 24579
    },
    {
      "epoch": 5.851514610486222,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 24580
    },
    {
      "epoch": 5.851752663214902,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 24581
    },
    {
      "epoch": 5.8519907159435816,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 24582
    },
    {
      "epoch": 5.852228768672261,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 24583
    },
    {
      "epoch": 5.85246682140094,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 24584
    },
    {
      "epoch": 5.8527048741296195,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 24585
    },
    {
      "epoch": 5.852942926858299,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2268,
      "step": 24586
    },
    {
      "epoch": 5.853180979586979,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3649,
      "step": 24587
    },
    {
      "epoch": 5.853419032315658,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 24588
    },
    {
      "epoch": 5.853657085044337,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 24589
    },
    {
      "epoch": 5.853895137773017,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3687,
      "step": 24590
    },
    {
      "epoch": 5.854133190501696,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 24591
    },
    {
      "epoch": 5.854371243230376,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 24592
    },
    {
      "epoch": 5.8546092959590545,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2326,
      "step": 24593
    },
    {
      "epoch": 5.854847348687734,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 24594
    },
    {
      "epoch": 5.855085401416414,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 24595
    },
    {
      "epoch": 5.855323454145093,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 24596
    },
    {
      "epoch": 5.855561506873773,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 24597
    },
    {
      "epoch": 5.855799559602452,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 24598
    },
    {
      "epoch": 5.856037612331131,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 24599
    },
    {
      "epoch": 5.856275665059811,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2287,
      "step": 24600
    },
    {
      "epoch": 5.856275665059811,
      "eval_runtime": 0.5343,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 1.872,
      "step": 24600
    },
    {
      "epoch": 5.8565137177884905,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 24601
    },
    {
      "epoch": 5.856751770517169,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3934,
      "step": 24602
    },
    {
      "epoch": 5.856989823245849,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 24603
    },
    {
      "epoch": 5.857227875974528,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.4092,
      "step": 24604
    },
    {
      "epoch": 5.857465928703208,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 24605
    },
    {
      "epoch": 5.857703981431888,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 24606
    },
    {
      "epoch": 5.857942034160566,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 24607
    },
    {
      "epoch": 5.858180086889246,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 24608
    },
    {
      "epoch": 5.858418139617926,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2364,
      "step": 24609
    },
    {
      "epoch": 5.858656192346604,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 24610
    },
    {
      "epoch": 5.858894245075284,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 24611
    },
    {
      "epoch": 5.8591322978039635,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 24612
    },
    {
      "epoch": 5.859370350532643,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9068,
      "step": 24613
    },
    {
      "epoch": 5.859608403261323,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 24614
    },
    {
      "epoch": 5.859846455990001,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 24615
    },
    {
      "epoch": 5.860084508718681,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2463,
      "step": 24616
    },
    {
      "epoch": 5.860322561447361,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3697,
      "step": 24617
    },
    {
      "epoch": 5.86056061417604,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4538,
      "step": 24618
    },
    {
      "epoch": 5.860798666904719,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 24619
    },
    {
      "epoch": 5.861036719633399,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 24620
    },
    {
      "epoch": 5.861274772362078,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 24621
    },
    {
      "epoch": 5.861512825090758,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 24622
    },
    {
      "epoch": 5.861750877819437,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 24623
    },
    {
      "epoch": 5.861988930548116,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 24624
    },
    {
      "epoch": 5.862226983276796,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 24625
    },
    {
      "epoch": 5.862465036005475,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 24626
    },
    {
      "epoch": 5.862703088734155,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 24627
    },
    {
      "epoch": 5.862941141462834,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 24628
    },
    {
      "epoch": 5.863179194191513,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 24629
    },
    {
      "epoch": 5.863417246920193,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 24630
    },
    {
      "epoch": 5.863655299648872,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3819,
      "step": 24631
    },
    {
      "epoch": 5.863893352377552,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 24632
    },
    {
      "epoch": 5.864131405106231,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 24633
    },
    {
      "epoch": 5.86436945783491,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 24634
    },
    {
      "epoch": 5.86460751056359,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 24635
    },
    {
      "epoch": 5.86484556329227,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 24636
    },
    {
      "epoch": 5.865083616020948,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 24637
    },
    {
      "epoch": 5.865321668749628,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 24638
    },
    {
      "epoch": 5.8655597214783075,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 24639
    },
    {
      "epoch": 5.865797774206987,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 24640
    },
    {
      "epoch": 5.866035826935667,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 24641
    },
    {
      "epoch": 5.866273879664345,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 24642
    },
    {
      "epoch": 5.866511932393025,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7352,
      "step": 24643
    },
    {
      "epoch": 5.866749985121705,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.348,
      "step": 24644
    },
    {
      "epoch": 5.866988037850383,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 24645
    },
    {
      "epoch": 5.867226090579063,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 24646
    },
    {
      "epoch": 5.867464143307743,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 24647
    },
    {
      "epoch": 5.867702196036422,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 24648
    },
    {
      "epoch": 5.867940248765102,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 24649
    },
    {
      "epoch": 5.8681783014937805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 24650
    },
    {
      "epoch": 5.86841635422246,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 24651
    },
    {
      "epoch": 5.86865440695114,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 24652
    },
    {
      "epoch": 5.868892459679819,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 24653
    },
    {
      "epoch": 5.869130512408498,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 24654
    },
    {
      "epoch": 5.869368565137178,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4161,
      "step": 24655
    },
    {
      "epoch": 5.869606617865857,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 24656
    },
    {
      "epoch": 5.869844670594537,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 24657
    },
    {
      "epoch": 5.8700827233232165,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 24658
    },
    {
      "epoch": 5.870320776051895,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8978,
      "step": 24659
    },
    {
      "epoch": 5.870558828780575,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 24660
    },
    {
      "epoch": 5.870796881509254,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 24661
    },
    {
      "epoch": 5.871034934237934,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2292,
      "step": 24662
    },
    {
      "epoch": 5.871272986966613,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 24663
    },
    {
      "epoch": 5.871511039695292,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 24664
    },
    {
      "epoch": 5.871749092423972,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0654,
      "step": 24665
    },
    {
      "epoch": 5.8719871451526515,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 24666
    },
    {
      "epoch": 5.872225197881331,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 24667
    },
    {
      "epoch": 5.87246325061001,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 24668
    },
    {
      "epoch": 5.8727013033386894,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 24669
    },
    {
      "epoch": 5.872939356067369,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 24670
    },
    {
      "epoch": 5.873177408796049,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3674,
      "step": 24671
    },
    {
      "epoch": 5.873415461524727,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 24672
    },
    {
      "epoch": 5.873653514253407,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 24673
    },
    {
      "epoch": 5.873891566982087,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 24674
    },
    {
      "epoch": 5.874129619710766,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 24675
    },
    {
      "epoch": 5.874367672439446,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 24676
    },
    {
      "epoch": 5.8746057251681245,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 24677
    },
    {
      "epoch": 5.874843777896804,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 24678
    },
    {
      "epoch": 5.875081830625484,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 24679
    },
    {
      "epoch": 5.875319883354163,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 24680
    },
    {
      "epoch": 5.875557936082842,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 24681
    },
    {
      "epoch": 5.875795988811522,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 24682
    },
    {
      "epoch": 5.876034041540201,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 24683
    },
    {
      "epoch": 5.876272094268881,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 24684
    },
    {
      "epoch": 5.87651014699756,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 24685
    },
    {
      "epoch": 5.876748199726239,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2947,
      "step": 24686
    },
    {
      "epoch": 5.876986252454919,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 24687
    },
    {
      "epoch": 5.877224305183598,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 24688
    },
    {
      "epoch": 5.877462357912277,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 24689
    },
    {
      "epoch": 5.877700410640957,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 24690
    },
    {
      "epoch": 5.877938463369636,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 24691
    },
    {
      "epoch": 5.878176516098316,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 24692
    },
    {
      "epoch": 5.8784145688269955,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 24693
    },
    {
      "epoch": 5.878652621555674,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 24694
    },
    {
      "epoch": 5.878890674284354,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 24695
    },
    {
      "epoch": 5.8791287270130335,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 24696
    },
    {
      "epoch": 5.879366779741713,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 24697
    },
    {
      "epoch": 5.879604832470392,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 24698
    },
    {
      "epoch": 5.879842885199071,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 24699
    },
    {
      "epoch": 5.880080937927751,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 24700
    },
    {
      "epoch": 5.880080937927751,
      "eval_runtime": 0.5344,
      "eval_samples_per_second": 1.871,
      "eval_steps_per_second": 1.871,
      "step": 24700
    },
    {
      "epoch": 5.880318990656431,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3144,
      "step": 24701
    },
    {
      "epoch": 5.88055704338511,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 24702
    },
    {
      "epoch": 5.880795096113789,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 24703
    },
    {
      "epoch": 5.8810331488424685,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.433,
      "step": 24704
    },
    {
      "epoch": 5.881271201571148,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 24705
    },
    {
      "epoch": 5.881509254299828,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 24706
    },
    {
      "epoch": 5.8817473070285065,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 24707
    },
    {
      "epoch": 5.881985359757186,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 24708
    },
    {
      "epoch": 5.882223412485866,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 24709
    },
    {
      "epoch": 5.882461465214545,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 24710
    },
    {
      "epoch": 5.882699517943225,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 24711
    },
    {
      "epoch": 5.882937570671904,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 24712
    },
    {
      "epoch": 5.883175623400583,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 24713
    },
    {
      "epoch": 5.883413676129263,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 24714
    },
    {
      "epoch": 5.883651728857942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 24715
    },
    {
      "epoch": 5.883889781586621,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3634,
      "step": 24716
    },
    {
      "epoch": 5.884127834315301,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 24717
    },
    {
      "epoch": 5.88436588704398,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 24718
    },
    {
      "epoch": 5.88460393977266,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 24719
    },
    {
      "epoch": 5.884841992501339,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 24720
    },
    {
      "epoch": 5.885080045230018,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 24721
    },
    {
      "epoch": 5.885318097958698,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 24722
    },
    {
      "epoch": 5.8855561506873775,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 24723
    },
    {
      "epoch": 5.885794203416056,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 24724
    },
    {
      "epoch": 5.886032256144736,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.7265,
      "step": 24725
    },
    {
      "epoch": 5.886270308873415,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 24726
    },
    {
      "epoch": 5.886508361602095,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 24727
    },
    {
      "epoch": 5.886746414330775,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 24728
    },
    {
      "epoch": 5.886984467059453,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 24729
    },
    {
      "epoch": 5.887222519788133,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 24730
    },
    {
      "epoch": 5.8874605725168125,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3003,
      "step": 24731
    },
    {
      "epoch": 5.887698625245492,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 24732
    },
    {
      "epoch": 5.887936677974171,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 24733
    },
    {
      "epoch": 5.8881747307028505,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 24734
    },
    {
      "epoch": 5.88841278343153,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3158,
      "step": 24735
    },
    {
      "epoch": 5.88865083616021,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 24736
    },
    {
      "epoch": 5.888888888888889,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 24737
    },
    {
      "epoch": 5.889126941617568,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.3806,
      "step": 24738
    },
    {
      "epoch": 5.889364994346248,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 24739
    },
    {
      "epoch": 5.889603047074927,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 24740
    },
    {
      "epoch": 5.889841099803607,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 24741
    },
    {
      "epoch": 5.8900791525322855,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 24742
    },
    {
      "epoch": 5.890317205260965,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 24743
    },
    {
      "epoch": 5.890555257989645,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1163,
      "step": 24744
    },
    {
      "epoch": 5.890793310718324,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 24745
    },
    {
      "epoch": 5.891031363447004,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 24746
    },
    {
      "epoch": 5.891269416175683,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 24747
    },
    {
      "epoch": 5.891507468904362,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 24748
    },
    {
      "epoch": 5.891745521633042,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 24749
    },
    {
      "epoch": 5.8919835743617215,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 24750
    },
    {
      "epoch": 5.8922216270904,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 24751
    },
    {
      "epoch": 5.89245967981908,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2763,
      "step": 24752
    },
    {
      "epoch": 5.892697732547759,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 24753
    },
    {
      "epoch": 5.892935785276439,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 24754
    },
    {
      "epoch": 5.893173838005119,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 24755
    },
    {
      "epoch": 5.893411890733797,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 24756
    },
    {
      "epoch": 5.893649943462477,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 24757
    },
    {
      "epoch": 5.8938879961911566,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 24758
    },
    {
      "epoch": 5.894126048919835,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 24759
    },
    {
      "epoch": 5.894364101648515,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7158,
      "step": 24760
    },
    {
      "epoch": 5.8946021543771945,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1843,
      "step": 24761
    },
    {
      "epoch": 5.894840207105874,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3992,
      "step": 24762
    },
    {
      "epoch": 5.895078259834554,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 24763
    },
    {
      "epoch": 5.895316312563232,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 24764
    },
    {
      "epoch": 5.895554365291912,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 24765
    },
    {
      "epoch": 5.895792418020592,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 24766
    },
    {
      "epoch": 5.896030470749271,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 24767
    },
    {
      "epoch": 5.89626852347795,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 24768
    },
    {
      "epoch": 5.8965065762066295,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 24769
    },
    {
      "epoch": 5.896744628935309,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 24770
    },
    {
      "epoch": 5.896982681663989,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3345,
      "step": 24771
    },
    {
      "epoch": 5.897220734392668,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 24772
    },
    {
      "epoch": 5.897458787121347,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 24773
    },
    {
      "epoch": 5.897696839850027,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 24774
    },
    {
      "epoch": 5.897934892578706,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.276,
      "step": 24775
    },
    {
      "epoch": 5.898172945307386,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9397,
      "step": 24776
    },
    {
      "epoch": 5.898410998036065,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 24777
    },
    {
      "epoch": 5.898649050764744,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 24778
    },
    {
      "epoch": 5.898887103493424,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 24779
    },
    {
      "epoch": 5.899125156222103,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 24780
    },
    {
      "epoch": 5.899363208950783,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3596,
      "step": 24781
    },
    {
      "epoch": 5.899601261679462,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 24782
    },
    {
      "epoch": 5.899839314408141,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 24783
    },
    {
      "epoch": 5.900077367136821,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 24784
    },
    {
      "epoch": 5.900315419865501,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 24785
    },
    {
      "epoch": 5.900553472594179,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 24786
    },
    {
      "epoch": 5.900791525322859,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 24787
    },
    {
      "epoch": 5.9010295780515385,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 24788
    },
    {
      "epoch": 5.901267630780218,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 24789
    },
    {
      "epoch": 5.901505683508898,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 24790
    },
    {
      "epoch": 5.901743736237576,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 24791
    },
    {
      "epoch": 5.901981788966256,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 24792
    },
    {
      "epoch": 5.902219841694936,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 24793
    },
    {
      "epoch": 5.902457894423614,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 24794
    },
    {
      "epoch": 5.902695947152294,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 24795
    },
    {
      "epoch": 5.902933999880974,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 24796
    },
    {
      "epoch": 5.903172052609653,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 24797
    },
    {
      "epoch": 5.903410105338333,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1633,
      "step": 24798
    },
    {
      "epoch": 5.9036481580670115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 24799
    },
    {
      "epoch": 5.903886210795691,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 24800
    },
    {
      "epoch": 5.903886210795691,
      "eval_runtime": 0.5315,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 24800
    },
    {
      "epoch": 5.904124263524371,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 24801
    },
    {
      "epoch": 5.90436231625305,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 24802
    },
    {
      "epoch": 5.904600368981729,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3574,
      "step": 24803
    },
    {
      "epoch": 5.904838421710409,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 24804
    },
    {
      "epoch": 5.905076474439088,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 24805
    },
    {
      "epoch": 5.905314527167768,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 24806
    },
    {
      "epoch": 5.905552579896447,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 24807
    },
    {
      "epoch": 5.905790632625126,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 24808
    },
    {
      "epoch": 5.906028685353806,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.4115,
      "step": 24809
    },
    {
      "epoch": 5.906266738082485,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 24810
    },
    {
      "epoch": 5.906504790811165,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 24811
    },
    {
      "epoch": 5.906742843539844,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 24812
    },
    {
      "epoch": 5.906980896268523,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4432,
      "step": 24813
    },
    {
      "epoch": 5.907218948997203,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 24814
    },
    {
      "epoch": 5.9074570017258825,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 24815
    },
    {
      "epoch": 5.907695054454562,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 24816
    },
    {
      "epoch": 5.907933107183241,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 24817
    },
    {
      "epoch": 5.90817115991192,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 24818
    },
    {
      "epoch": 5.9084092126406,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 24819
    },
    {
      "epoch": 5.90864726536928,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 24820
    },
    {
      "epoch": 5.908885318097958,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 24821
    },
    {
      "epoch": 5.909123370826638,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2496,
      "step": 24822
    },
    {
      "epoch": 5.909361423555318,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 24823
    },
    {
      "epoch": 5.909599476283997,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 24824
    },
    {
      "epoch": 5.909837529012677,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 24825
    },
    {
      "epoch": 5.9100755817413555,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 24826
    },
    {
      "epoch": 5.910313634470035,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 24827
    },
    {
      "epoch": 5.910551687198715,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 24828
    },
    {
      "epoch": 5.910789739927393,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5359,
      "step": 24829
    },
    {
      "epoch": 5.911027792656073,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 24830
    },
    {
      "epoch": 5.911265845384753,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 24831
    },
    {
      "epoch": 5.911503898113432,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8389,
      "step": 24832
    },
    {
      "epoch": 5.911741950842112,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 24833
    },
    {
      "epoch": 5.911980003570791,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 24834
    },
    {
      "epoch": 5.91221805629947,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 24835
    },
    {
      "epoch": 5.91245610902815,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 24836
    },
    {
      "epoch": 5.912694161756829,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 24837
    },
    {
      "epoch": 5.912932214485508,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 24838
    },
    {
      "epoch": 5.913170267214188,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3832,
      "step": 24839
    },
    {
      "epoch": 5.913408319942867,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3223,
      "step": 24840
    },
    {
      "epoch": 5.913646372671547,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 24841
    },
    {
      "epoch": 5.9138844254002265,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 24842
    },
    {
      "epoch": 5.914122478128905,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 24843
    },
    {
      "epoch": 5.914360530857585,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 24844
    },
    {
      "epoch": 5.9145985835862644,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 24845
    },
    {
      "epoch": 5.914836636314944,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 24846
    },
    {
      "epoch": 5.915074689043623,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.565,
      "step": 24847
    },
    {
      "epoch": 5.915312741772302,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 24848
    },
    {
      "epoch": 5.915550794500982,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 24849
    },
    {
      "epoch": 5.915788847229662,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4538,
      "step": 24850
    },
    {
      "epoch": 5.916026899958341,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 24851
    },
    {
      "epoch": 5.91626495268702,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 24852
    },
    {
      "epoch": 5.9165030054156995,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 24853
    },
    {
      "epoch": 5.916741058144379,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 24854
    },
    {
      "epoch": 5.916979110873059,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 24855
    },
    {
      "epoch": 5.917217163601737,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 24856
    },
    {
      "epoch": 5.917455216330417,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 24857
    },
    {
      "epoch": 5.917693269059097,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 24858
    },
    {
      "epoch": 5.917931321787776,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 24859
    },
    {
      "epoch": 5.918169374516456,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 24860
    },
    {
      "epoch": 5.918407427245135,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 24861
    },
    {
      "epoch": 5.918645479973814,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 24862
    },
    {
      "epoch": 5.918883532702494,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 24863
    },
    {
      "epoch": 5.9191215854311725,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 24864
    },
    {
      "epoch": 5.919359638159852,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 24865
    },
    {
      "epoch": 5.919597690888532,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 24866
    },
    {
      "epoch": 5.919835743617211,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 24867
    },
    {
      "epoch": 5.920073796345891,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 24868
    },
    {
      "epoch": 5.92031184907457,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 24869
    },
    {
      "epoch": 5.920549901803249,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 24870
    },
    {
      "epoch": 5.920787954531929,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 24871
    },
    {
      "epoch": 5.9210260072606085,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 24872
    },
    {
      "epoch": 5.921264059989287,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 24873
    },
    {
      "epoch": 5.921502112717967,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 24874
    },
    {
      "epoch": 5.921740165446646,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 24875
    },
    {
      "epoch": 5.921978218175326,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 24876
    },
    {
      "epoch": 5.922216270904006,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 24877
    },
    {
      "epoch": 5.922454323632684,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 24878
    },
    {
      "epoch": 5.922692376361364,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 24879
    },
    {
      "epoch": 5.9229304290900435,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 24880
    },
    {
      "epoch": 5.923168481818723,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 24881
    },
    {
      "epoch": 5.923406534547402,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 24882
    },
    {
      "epoch": 5.9236445872760815,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2176,
      "step": 24883
    },
    {
      "epoch": 5.923882640004761,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2478,
      "step": 24884
    },
    {
      "epoch": 5.924120692733441,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 24885
    },
    {
      "epoch": 5.92435874546212,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.7099,
      "step": 24886
    },
    {
      "epoch": 5.924596798190799,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 24887
    },
    {
      "epoch": 5.924834850919479,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 24888
    },
    {
      "epoch": 5.925072903648158,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 24889
    },
    {
      "epoch": 5.925310956376838,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 24890
    },
    {
      "epoch": 5.9255490091055165,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 24891
    },
    {
      "epoch": 5.925787061834196,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 24892
    },
    {
      "epoch": 5.926025114562876,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 24893
    },
    {
      "epoch": 5.926263167291555,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 24894
    },
    {
      "epoch": 5.926501220020235,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 24895
    },
    {
      "epoch": 5.926739272748914,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 24896
    },
    {
      "epoch": 5.926977325477593,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 24897
    },
    {
      "epoch": 5.927215378206273,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 24898
    },
    {
      "epoch": 5.9274534309349525,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 24899
    },
    {
      "epoch": 5.927691483663631,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 24900
    },
    {
      "epoch": 5.927691483663631,
      "eval_runtime": 0.5252,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 24900
    },
    {
      "epoch": 5.927929536392311,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 24901
    },
    {
      "epoch": 5.92816758912099,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 24902
    },
    {
      "epoch": 5.92840564184967,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4554,
      "step": 24903
    },
    {
      "epoch": 5.928643694578349,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3687,
      "step": 24904
    },
    {
      "epoch": 5.928881747307028,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 24905
    },
    {
      "epoch": 5.929119800035708,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 24906
    },
    {
      "epoch": 5.9293578527643875,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 24907
    },
    {
      "epoch": 5.929595905493066,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 24908
    },
    {
      "epoch": 5.929833958221746,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.469,
      "step": 24909
    },
    {
      "epoch": 5.9300720109504255,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 24910
    },
    {
      "epoch": 5.930310063679105,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 24911
    },
    {
      "epoch": 5.930548116407785,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 24912
    },
    {
      "epoch": 5.930786169136463,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 24913
    },
    {
      "epoch": 5.931024221865143,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 24914
    },
    {
      "epoch": 5.931262274593823,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.25,
      "step": 24915
    },
    {
      "epoch": 5.931500327322502,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 24916
    },
    {
      "epoch": 5.931738380051181,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 24917
    },
    {
      "epoch": 5.9319764327798605,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 24918
    },
    {
      "epoch": 5.93221448550854,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 24919
    },
    {
      "epoch": 5.93245253823722,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 24920
    },
    {
      "epoch": 5.932690590965899,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 24921
    },
    {
      "epoch": 5.932928643694578,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8342,
      "step": 24922
    },
    {
      "epoch": 5.933166696423258,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 24923
    },
    {
      "epoch": 5.933404749151937,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 24924
    },
    {
      "epoch": 5.933642801880617,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 24925
    },
    {
      "epoch": 5.933880854609296,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 24926
    },
    {
      "epoch": 5.934118907337975,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 24927
    },
    {
      "epoch": 5.934356960066655,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8762,
      "step": 24928
    },
    {
      "epoch": 5.934595012795334,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 24929
    },
    {
      "epoch": 5.934833065524014,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2895,
      "step": 24930
    },
    {
      "epoch": 5.935071118252693,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 24931
    },
    {
      "epoch": 5.935309170981372,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 24932
    },
    {
      "epoch": 5.935547223710052,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 24933
    },
    {
      "epoch": 5.9357852764387316,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 24934
    },
    {
      "epoch": 5.93602332916741,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0472,
      "step": 24935
    },
    {
      "epoch": 5.93626138189609,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 24936
    },
    {
      "epoch": 5.9364994346247695,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 24937
    },
    {
      "epoch": 5.936737487353449,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9172,
      "step": 24938
    },
    {
      "epoch": 5.936975540082129,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 24939
    },
    {
      "epoch": 5.937213592810807,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 24940
    },
    {
      "epoch": 5.937451645539487,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 24941
    },
    {
      "epoch": 5.937689698268167,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 24942
    },
    {
      "epoch": 5.937927750996845,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 24943
    },
    {
      "epoch": 5.938165803725525,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2798,
      "step": 24944
    },
    {
      "epoch": 5.9384038564542045,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 24945
    },
    {
      "epoch": 5.938641909182884,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2724,
      "step": 24946
    },
    {
      "epoch": 5.938879961911564,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 24947
    },
    {
      "epoch": 5.9391180146402425,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4218,
      "step": 24948
    },
    {
      "epoch": 5.939356067368922,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 24949
    },
    {
      "epoch": 5.939594120097602,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 24950
    },
    {
      "epoch": 5.939832172826281,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 24951
    },
    {
      "epoch": 5.94007022555496,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 24952
    },
    {
      "epoch": 5.94030827828364,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 24953
    },
    {
      "epoch": 5.940546331012319,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 24954
    },
    {
      "epoch": 5.940784383740999,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 24955
    },
    {
      "epoch": 5.941022436469678,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 24956
    },
    {
      "epoch": 5.941260489198357,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 24957
    },
    {
      "epoch": 5.941498541927037,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 24958
    },
    {
      "epoch": 5.941736594655716,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 24959
    },
    {
      "epoch": 5.941974647384396,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 24960
    },
    {
      "epoch": 5.942212700113075,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 24961
    },
    {
      "epoch": 5.942450752841754,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 24962
    },
    {
      "epoch": 5.942688805570434,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 24963
    },
    {
      "epoch": 5.9429268582991135,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2517,
      "step": 24964
    },
    {
      "epoch": 5.943164911027793,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 24965
    },
    {
      "epoch": 5.943402963756472,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 24966
    },
    {
      "epoch": 5.943641016485151,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 24967
    },
    {
      "epoch": 5.943879069213831,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 24968
    },
    {
      "epoch": 5.944117121942511,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 24969
    },
    {
      "epoch": 5.944355174671189,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 24970
    },
    {
      "epoch": 5.944593227399869,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 24971
    },
    {
      "epoch": 5.944831280128549,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 24972
    },
    {
      "epoch": 5.945069332857228,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 24973
    },
    {
      "epoch": 5.945307385585908,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 24974
    },
    {
      "epoch": 5.9455454383145865,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3262,
      "step": 24975
    },
    {
      "epoch": 5.945783491043266,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 24976
    },
    {
      "epoch": 5.946021543771946,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.207,
      "step": 24977
    },
    {
      "epoch": 5.946259596500624,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.409,
      "step": 24978
    },
    {
      "epoch": 5.946497649229304,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 24979
    },
    {
      "epoch": 5.946735701957984,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 24980
    },
    {
      "epoch": 5.946973754686663,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 24981
    },
    {
      "epoch": 5.947211807415343,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.7762,
      "step": 24982
    },
    {
      "epoch": 5.9474498601440215,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 24983
    },
    {
      "epoch": 5.947687912872701,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2876,
      "step": 24984
    },
    {
      "epoch": 5.947925965601381,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 24985
    },
    {
      "epoch": 5.94816401833006,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 24986
    },
    {
      "epoch": 5.948402071058739,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 24987
    },
    {
      "epoch": 5.948640123787419,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 24988
    },
    {
      "epoch": 5.948878176516098,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 24989
    },
    {
      "epoch": 5.949116229244778,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 24990
    },
    {
      "epoch": 5.9493542819734575,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4214,
      "step": 24991
    },
    {
      "epoch": 5.949592334702136,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 24992
    },
    {
      "epoch": 5.949830387430816,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 24993
    },
    {
      "epoch": 5.950068440159495,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 24994
    },
    {
      "epoch": 5.950306492888175,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 24995
    },
    {
      "epoch": 5.950544545616854,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 24996
    },
    {
      "epoch": 5.950782598345533,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 24997
    },
    {
      "epoch": 5.951020651074213,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 24998
    },
    {
      "epoch": 5.951258703802893,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3709,
      "step": 24999
    },
    {
      "epoch": 5.951496756531572,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 25000
    },
    {
      "epoch": 5.951496756531572,
      "eval_runtime": 0.5291,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 25000
    },
    {
      "epoch": 5.951734809260251,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3276,
      "step": 25001
    },
    {
      "epoch": 5.9519728619889305,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 25002
    },
    {
      "epoch": 5.95221091471761,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 25003
    },
    {
      "epoch": 5.95244896744629,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 25004
    },
    {
      "epoch": 5.952687020174968,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 25005
    },
    {
      "epoch": 5.952925072903648,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 25006
    },
    {
      "epoch": 5.953163125632328,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 25007
    },
    {
      "epoch": 5.953401178361007,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.268,
      "step": 25008
    },
    {
      "epoch": 5.953639231089687,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 25009
    },
    {
      "epoch": 5.953877283818366,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3118,
      "step": 25010
    },
    {
      "epoch": 5.954115336547045,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 25011
    },
    {
      "epoch": 5.954353389275725,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 25012
    },
    {
      "epoch": 5.9545914420044035,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 25013
    },
    {
      "epoch": 5.954829494733083,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 25014
    },
    {
      "epoch": 5.955067547461763,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 25015
    },
    {
      "epoch": 5.955305600190442,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 25016
    },
    {
      "epoch": 5.955543652919122,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 25017
    },
    {
      "epoch": 5.955781705647801,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.6738,
      "step": 25018
    },
    {
      "epoch": 5.95601975837648,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 25019
    },
    {
      "epoch": 5.95625781110516,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 25020
    },
    {
      "epoch": 5.956495863833839,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.5622,
      "step": 25021
    },
    {
      "epoch": 5.956733916562518,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 25022
    },
    {
      "epoch": 5.956971969291198,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 25023
    },
    {
      "epoch": 5.957210022019877,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8178,
      "step": 25024
    },
    {
      "epoch": 5.957448074748557,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3001,
      "step": 25025
    },
    {
      "epoch": 5.957686127477237,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 25026
    },
    {
      "epoch": 5.957924180205915,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 25027
    },
    {
      "epoch": 5.958162232934595,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3954,
      "step": 25028
    },
    {
      "epoch": 5.9584002856632745,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 25029
    },
    {
      "epoch": 5.958638338391954,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 25030
    },
    {
      "epoch": 5.958876391120633,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 25031
    },
    {
      "epoch": 5.959114443849312,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 25032
    },
    {
      "epoch": 5.959352496577992,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 25033
    },
    {
      "epoch": 5.959590549306672,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 25034
    },
    {
      "epoch": 5.959828602035351,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 25035
    },
    {
      "epoch": 5.96006665476403,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 25036
    },
    {
      "epoch": 5.96030470749271,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 25037
    },
    {
      "epoch": 5.960542760221389,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.6844,
      "step": 25038
    },
    {
      "epoch": 5.960780812950069,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 25039
    },
    {
      "epoch": 5.9610188656787475,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 25040
    },
    {
      "epoch": 5.961256918407427,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 25041
    },
    {
      "epoch": 5.961494971136107,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2501,
      "step": 25042
    },
    {
      "epoch": 5.961733023864786,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1057,
      "step": 25043
    },
    {
      "epoch": 5.961971076593466,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 25044
    },
    {
      "epoch": 5.962209129322145,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 25045
    },
    {
      "epoch": 5.962447182050824,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5387,
      "step": 25046
    },
    {
      "epoch": 5.962685234779504,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4607,
      "step": 25047
    },
    {
      "epoch": 5.962923287508183,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 25048
    },
    {
      "epoch": 5.963161340236862,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 25049
    },
    {
      "epoch": 5.963399392965542,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2812,
      "step": 25050
    },
    {
      "epoch": 5.963637445694221,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 25051
    },
    {
      "epoch": 5.963875498422901,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 25052
    },
    {
      "epoch": 5.96411355115158,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3913,
      "step": 25053
    },
    {
      "epoch": 5.964351603880259,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 25054
    },
    {
      "epoch": 5.964589656608939,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 25055
    },
    {
      "epoch": 5.9648277093376185,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.3691,
      "step": 25056
    },
    {
      "epoch": 5.965065762066297,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.114,
      "step": 25057
    },
    {
      "epoch": 5.965303814794977,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 25058
    },
    {
      "epoch": 5.9655418675236564,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 25059
    },
    {
      "epoch": 5.965779920252336,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0677,
      "step": 25060
    },
    {
      "epoch": 5.966017972981016,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 25061
    },
    {
      "epoch": 5.966256025709694,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 25062
    },
    {
      "epoch": 5.966494078438374,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 25063
    },
    {
      "epoch": 5.966732131167054,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 25064
    },
    {
      "epoch": 5.966970183895733,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 25065
    },
    {
      "epoch": 5.967208236624412,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 25066
    },
    {
      "epoch": 5.9674462893530915,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2246,
      "step": 25067
    },
    {
      "epoch": 5.967684342081771,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 25068
    },
    {
      "epoch": 5.967922394810451,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 25069
    },
    {
      "epoch": 5.96816044753913,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 25070
    },
    {
      "epoch": 5.968398500267809,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 25071
    },
    {
      "epoch": 5.968636552996489,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 25072
    },
    {
      "epoch": 5.968874605725168,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 25073
    },
    {
      "epoch": 5.969112658453848,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 25074
    },
    {
      "epoch": 5.969350711182527,
      "grad_norm": 6.1875,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 25075
    },
    {
      "epoch": 5.969588763911206,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1014,
      "step": 25076
    },
    {
      "epoch": 5.969826816639886,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 25077
    },
    {
      "epoch": 5.970064869368565,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 25078
    },
    {
      "epoch": 5.970302922097245,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 25079
    },
    {
      "epoch": 5.970540974825924,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 25080
    },
    {
      "epoch": 5.970779027554603,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 25081
    },
    {
      "epoch": 5.971017080283283,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 25082
    },
    {
      "epoch": 5.9712551330119625,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 25083
    },
    {
      "epoch": 5.971493185740641,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 25084
    },
    {
      "epoch": 5.971731238469321,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 25085
    },
    {
      "epoch": 5.9719692911980005,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2689,
      "step": 25086
    },
    {
      "epoch": 5.97220734392668,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 25087
    },
    {
      "epoch": 5.972445396655359,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 25088
    },
    {
      "epoch": 5.972683449384038,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 25089
    },
    {
      "epoch": 5.972921502112718,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 25090
    },
    {
      "epoch": 5.973159554841398,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 25091
    },
    {
      "epoch": 5.973397607570076,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 25092
    },
    {
      "epoch": 5.973635660298756,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 25093
    },
    {
      "epoch": 5.9738737130274355,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 25094
    },
    {
      "epoch": 5.974111765756115,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 25095
    },
    {
      "epoch": 5.974349818484795,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 25096
    },
    {
      "epoch": 5.9745878712134735,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 25097
    },
    {
      "epoch": 5.974825923942153,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6176,
      "step": 25098
    },
    {
      "epoch": 5.975063976670833,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 25099
    },
    {
      "epoch": 5.975302029399512,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3839,
      "step": 25100
    },
    {
      "epoch": 5.975302029399512,
      "eval_runtime": 0.5298,
      "eval_samples_per_second": 1.888,
      "eval_steps_per_second": 1.888,
      "step": 25100
    },
    {
      "epoch": 5.975540082128191,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 25101
    },
    {
      "epoch": 5.975778134856871,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 25102
    },
    {
      "epoch": 5.97601618758555,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 25103
    },
    {
      "epoch": 5.97625424031423,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8379,
      "step": 25104
    },
    {
      "epoch": 5.976492293042909,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 25105
    },
    {
      "epoch": 5.976730345771588,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.315,
      "step": 25106
    },
    {
      "epoch": 5.976968398500268,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 25107
    },
    {
      "epoch": 5.977206451228947,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 25108
    },
    {
      "epoch": 5.977444503957627,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 25109
    },
    {
      "epoch": 5.977682556686306,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4271,
      "step": 25110
    },
    {
      "epoch": 5.977920609414985,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 25111
    },
    {
      "epoch": 5.978158662143665,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 25112
    },
    {
      "epoch": 5.9783967148723445,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 25113
    },
    {
      "epoch": 5.978634767601024,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 25114
    },
    {
      "epoch": 5.978872820329703,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 25115
    },
    {
      "epoch": 5.979110873058382,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 25116
    },
    {
      "epoch": 5.979348925787062,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1614,
      "step": 25117
    },
    {
      "epoch": 5.979586978515742,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 25118
    },
    {
      "epoch": 5.97982503124442,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 25119
    },
    {
      "epoch": 5.9800630839731,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 25120
    },
    {
      "epoch": 5.9803011367017795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 25121
    },
    {
      "epoch": 5.980539189430459,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4856,
      "step": 25122
    },
    {
      "epoch": 5.980777242159138,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 25123
    },
    {
      "epoch": 5.9810152948878175,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 25124
    },
    {
      "epoch": 5.981253347616497,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 25125
    },
    {
      "epoch": 5.981491400345177,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4855,
      "step": 25126
    },
    {
      "epoch": 5.981729453073855,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 25127
    },
    {
      "epoch": 5.981967505802535,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 25128
    },
    {
      "epoch": 5.982205558531215,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 25129
    },
    {
      "epoch": 5.982443611259894,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 25130
    },
    {
      "epoch": 5.982681663988574,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 25131
    },
    {
      "epoch": 5.9829197167172525,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 25132
    },
    {
      "epoch": 5.983157769445932,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4015,
      "step": 25133
    },
    {
      "epoch": 5.983395822174612,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 25134
    },
    {
      "epoch": 5.983633874903291,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 25135
    },
    {
      "epoch": 5.98387192763197,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 25136
    },
    {
      "epoch": 5.98410998036065,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 25137
    },
    {
      "epoch": 5.984348033089329,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 25138
    },
    {
      "epoch": 5.984586085818009,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 25139
    },
    {
      "epoch": 5.9848241385466885,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 25140
    },
    {
      "epoch": 5.985062191275367,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 25141
    },
    {
      "epoch": 5.985300244004047,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 25142
    },
    {
      "epoch": 5.985538296732726,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8009,
      "step": 25143
    },
    {
      "epoch": 5.985776349461406,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 25144
    },
    {
      "epoch": 5.986014402190085,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 25145
    },
    {
      "epoch": 5.986252454918764,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 25146
    },
    {
      "epoch": 5.986490507647444,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 25147
    },
    {
      "epoch": 5.9867285603761236,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 25148
    },
    {
      "epoch": 5.986966613104803,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 25149
    },
    {
      "epoch": 5.987204665833482,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 25150
    },
    {
      "epoch": 5.9874427185621615,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 25151
    },
    {
      "epoch": 5.987680771290841,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 25152
    },
    {
      "epoch": 5.987918824019521,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 25153
    },
    {
      "epoch": 5.988156876748199,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3747,
      "step": 25154
    },
    {
      "epoch": 5.988394929476879,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 25155
    },
    {
      "epoch": 5.988632982205559,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 25156
    },
    {
      "epoch": 5.988871034934238,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 25157
    },
    {
      "epoch": 5.989109087662918,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 25158
    },
    {
      "epoch": 5.9893471403915965,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 25159
    },
    {
      "epoch": 5.989585193120276,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 25160
    },
    {
      "epoch": 5.989823245848956,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 25161
    },
    {
      "epoch": 5.9900612985776345,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 25162
    },
    {
      "epoch": 5.990299351306314,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 25163
    },
    {
      "epoch": 5.990537404034994,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8358,
      "step": 25164
    },
    {
      "epoch": 5.990775456763673,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 25165
    },
    {
      "epoch": 5.991013509492353,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3152,
      "step": 25166
    },
    {
      "epoch": 5.991251562221032,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 25167
    },
    {
      "epoch": 5.991489614949711,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3719,
      "step": 25168
    },
    {
      "epoch": 5.991727667678391,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3928,
      "step": 25169
    },
    {
      "epoch": 5.99196572040707,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 25170
    },
    {
      "epoch": 5.992203773135749,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 25171
    },
    {
      "epoch": 5.992441825864429,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 25172
    },
    {
      "epoch": 5.992679878593108,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 25173
    },
    {
      "epoch": 5.992917931321788,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.866,
      "step": 25174
    },
    {
      "epoch": 5.993155984050468,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4359,
      "step": 25175
    },
    {
      "epoch": 5.993394036779146,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1086,
      "step": 25176
    },
    {
      "epoch": 5.993632089507826,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 25177
    },
    {
      "epoch": 5.9938701422365055,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4258,
      "step": 25178
    },
    {
      "epoch": 5.994108194965185,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 25179
    },
    {
      "epoch": 5.994346247693864,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 25180
    },
    {
      "epoch": 5.994584300422543,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3317,
      "step": 25181
    },
    {
      "epoch": 5.994822353151223,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 25182
    },
    {
      "epoch": 5.995060405879903,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 25183
    },
    {
      "epoch": 5.995298458608582,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 25184
    },
    {
      "epoch": 5.995536511337261,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 25185
    },
    {
      "epoch": 5.995774564065941,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 25186
    },
    {
      "epoch": 5.99601261679462,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 25187
    },
    {
      "epoch": 5.9962506695233,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 25188
    },
    {
      "epoch": 5.9964887222519785,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 25189
    },
    {
      "epoch": 5.996726774980658,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 25190
    },
    {
      "epoch": 5.996964827709338,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 25191
    },
    {
      "epoch": 5.997202880438017,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 25192
    },
    {
      "epoch": 5.997440933166697,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 25193
    },
    {
      "epoch": 5.997678985895376,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 25194
    },
    {
      "epoch": 5.997917038624055,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 25195
    },
    {
      "epoch": 5.998155091352735,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 25196
    },
    {
      "epoch": 5.9983931440814136,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 25197
    },
    {
      "epoch": 5.998631196810093,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 25198
    },
    {
      "epoch": 5.998869249538773,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 25199
    },
    {
      "epoch": 5.999107302267452,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 25200
    },
    {
      "epoch": 5.999107302267452,
      "eval_runtime": 0.5198,
      "eval_samples_per_second": 1.924,
      "eval_steps_per_second": 1.924,
      "step": 25200
    },
    {
      "epoch": 5.999345354996132,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 25201
    },
    {
      "epoch": 5.999583407724811,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 25202
    },
    {
      "epoch": 5.99982146045349,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 25203
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5464,
      "step": 25204
    },
    {
      "epoch": 6.00023805272868,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 25205
    },
    {
      "epoch": 6.000476105457359,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 25206
    },
    {
      "epoch": 6.000714158186038,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 25207
    },
    {
      "epoch": 6.0009522109147175,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 25208
    },
    {
      "epoch": 6.001190263643397,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 25209
    },
    {
      "epoch": 6.001428316372077,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.951,
      "step": 25210
    },
    {
      "epoch": 6.0016663691007555,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 25211
    },
    {
      "epoch": 6.001904421829435,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 25212
    },
    {
      "epoch": 6.002142474558115,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 25213
    },
    {
      "epoch": 6.002380527286794,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 25214
    },
    {
      "epoch": 6.002618580015473,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 25215
    },
    {
      "epoch": 6.002856632744153,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 25216
    },
    {
      "epoch": 6.003094685472832,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 25217
    },
    {
      "epoch": 6.003332738201512,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 25218
    },
    {
      "epoch": 6.003570790930191,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 25219
    },
    {
      "epoch": 6.00380884365887,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 25220
    },
    {
      "epoch": 6.00404689638755,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 25221
    },
    {
      "epoch": 6.004284949116229,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 25222
    },
    {
      "epoch": 6.004523001844909,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.207,
      "step": 25223
    },
    {
      "epoch": 6.004761054573588,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 25224
    },
    {
      "epoch": 6.004999107302267,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 25225
    },
    {
      "epoch": 6.005237160030947,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 25226
    },
    {
      "epoch": 6.0054752127596265,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3544,
      "step": 25227
    },
    {
      "epoch": 6.005713265488306,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 25228
    },
    {
      "epoch": 6.005951318216985,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 25229
    },
    {
      "epoch": 6.006189370945664,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 25230
    },
    {
      "epoch": 6.006427423674344,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 25231
    },
    {
      "epoch": 6.006665476403024,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 25232
    },
    {
      "epoch": 6.006903529131702,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 25233
    },
    {
      "epoch": 6.007141581860382,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 25234
    },
    {
      "epoch": 6.0073796345890615,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 25235
    },
    {
      "epoch": 6.007617687317741,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 25236
    },
    {
      "epoch": 6.00785574004642,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 25237
    },
    {
      "epoch": 6.0080937927750995,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 25238
    },
    {
      "epoch": 6.008331845503779,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3836,
      "step": 25239
    },
    {
      "epoch": 6.008569898232459,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 25240
    },
    {
      "epoch": 6.008807950961138,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 25241
    },
    {
      "epoch": 6.009046003689817,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 25242
    },
    {
      "epoch": 6.009284056418497,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 25243
    },
    {
      "epoch": 6.009522109147176,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 25244
    },
    {
      "epoch": 6.009760161875856,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 25245
    },
    {
      "epoch": 6.0099982146045345,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3169,
      "step": 25246
    },
    {
      "epoch": 6.010236267333214,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 25247
    },
    {
      "epoch": 6.010474320061894,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 25248
    },
    {
      "epoch": 6.010712372790573,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 25249
    },
    {
      "epoch": 6.010950425519252,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3766,
      "step": 25250
    },
    {
      "epoch": 6.011188478247932,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 25251
    },
    {
      "epoch": 6.011426530976611,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 25252
    },
    {
      "epoch": 6.011664583705291,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2774,
      "step": 25253
    },
    {
      "epoch": 6.0119026364339705,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 25254
    },
    {
      "epoch": 6.012140689162649,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 25255
    },
    {
      "epoch": 6.012378741891329,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7617,
      "step": 25256
    },
    {
      "epoch": 6.012616794620008,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 25257
    },
    {
      "epoch": 6.012854847348688,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2196,
      "step": 25258
    },
    {
      "epoch": 6.013092900077367,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8398,
      "step": 25259
    },
    {
      "epoch": 6.013330952806046,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 25260
    },
    {
      "epoch": 6.013569005534726,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 25261
    },
    {
      "epoch": 6.013807058263406,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 25262
    },
    {
      "epoch": 6.014045110992085,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8393,
      "step": 25263
    },
    {
      "epoch": 6.014283163720764,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 25264
    },
    {
      "epoch": 6.0145212164494435,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2938,
      "step": 25265
    },
    {
      "epoch": 6.014759269178123,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 25266
    },
    {
      "epoch": 6.014997321906803,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 25267
    },
    {
      "epoch": 6.015235374635481,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 25268
    },
    {
      "epoch": 6.015473427364161,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9984,
      "step": 25269
    },
    {
      "epoch": 6.015711480092841,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9932,
      "step": 25270
    },
    {
      "epoch": 6.01594953282152,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 25271
    },
    {
      "epoch": 6.016187585550199,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 25272
    },
    {
      "epoch": 6.0164256382788786,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 25273
    },
    {
      "epoch": 6.016663691007558,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8956,
      "step": 25274
    },
    {
      "epoch": 6.016901743736238,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 25275
    },
    {
      "epoch": 6.017139796464917,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7913,
      "step": 25276
    },
    {
      "epoch": 6.017377849193596,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 25277
    },
    {
      "epoch": 6.017615901922276,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 25278
    },
    {
      "epoch": 6.017853954650955,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.5021,
      "step": 25279
    },
    {
      "epoch": 6.018092007379635,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 25280
    },
    {
      "epoch": 6.018330060108314,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 25281
    },
    {
      "epoch": 6.018568112836993,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 25282
    },
    {
      "epoch": 6.018806165565673,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 25283
    },
    {
      "epoch": 6.019044218294352,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 25284
    },
    {
      "epoch": 6.019282271023032,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 25285
    },
    {
      "epoch": 6.019520323751711,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 25286
    },
    {
      "epoch": 6.01975837648039,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 25287
    },
    {
      "epoch": 6.01999642920907,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 25288
    },
    {
      "epoch": 6.02023448193775,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 25289
    },
    {
      "epoch": 6.020472534666428,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 25290
    },
    {
      "epoch": 6.020710587395108,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4362,
      "step": 25291
    },
    {
      "epoch": 6.0209486401237875,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 25292
    },
    {
      "epoch": 6.021186692852467,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3569,
      "step": 25293
    },
    {
      "epoch": 6.021424745581146,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 25294
    },
    {
      "epoch": 6.021662798309825,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 25295
    },
    {
      "epoch": 6.021900851038505,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 25296
    },
    {
      "epoch": 6.022138903767185,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.295,
      "step": 25297
    },
    {
      "epoch": 6.022376956495864,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 25298
    },
    {
      "epoch": 6.022615009224543,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 25299
    },
    {
      "epoch": 6.022853061953223,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 25300
    },
    {
      "epoch": 6.022853061953223,
      "eval_runtime": 0.5241,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 25300
    },
    {
      "epoch": 6.023091114681902,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 25301
    },
    {
      "epoch": 6.023329167410582,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 25302
    },
    {
      "epoch": 6.0235672201392605,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 25303
    },
    {
      "epoch": 6.02380527286794,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 25304
    },
    {
      "epoch": 6.02404332559662,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 25305
    },
    {
      "epoch": 6.024281378325299,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2285,
      "step": 25306
    },
    {
      "epoch": 6.024519431053978,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 25307
    },
    {
      "epoch": 6.024757483782658,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 25308
    },
    {
      "epoch": 6.024995536511337,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 25309
    },
    {
      "epoch": 6.025233589240017,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3409,
      "step": 25310
    },
    {
      "epoch": 6.0254716419686964,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 25311
    },
    {
      "epoch": 6.025709694697375,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 25312
    },
    {
      "epoch": 6.025947747426055,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 25313
    },
    {
      "epoch": 6.026185800154734,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 25314
    },
    {
      "epoch": 6.026423852883414,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 25315
    },
    {
      "epoch": 6.026661905612093,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 25316
    },
    {
      "epoch": 6.026899958340772,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 25317
    },
    {
      "epoch": 6.027138011069452,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.856,
      "step": 25318
    },
    {
      "epoch": 6.0273760637981315,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 25319
    },
    {
      "epoch": 6.027614116526811,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 25320
    },
    {
      "epoch": 6.02785216925549,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 25321
    },
    {
      "epoch": 6.028090221984169,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 25322
    },
    {
      "epoch": 6.028328274712849,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 25323
    },
    {
      "epoch": 6.028566327441529,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4587,
      "step": 25324
    },
    {
      "epoch": 6.028804380170207,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 25325
    },
    {
      "epoch": 6.029042432898887,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 25326
    },
    {
      "epoch": 6.029280485627567,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 25327
    },
    {
      "epoch": 6.029518538356246,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 25328
    },
    {
      "epoch": 6.029756591084925,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 25329
    },
    {
      "epoch": 6.0299946438136045,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 25330
    },
    {
      "epoch": 6.030232696542284,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 25331
    },
    {
      "epoch": 6.030470749270964,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2726,
      "step": 25332
    },
    {
      "epoch": 6.030708801999643,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 25333
    },
    {
      "epoch": 6.030946854728322,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3528,
      "step": 25334
    },
    {
      "epoch": 6.031184907457002,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0891,
      "step": 25335
    },
    {
      "epoch": 6.031422960185681,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 25336
    },
    {
      "epoch": 6.031661012914361,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 25337
    },
    {
      "epoch": 6.03189906564304,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 25338
    },
    {
      "epoch": 6.032137118371719,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 25339
    },
    {
      "epoch": 6.032375171100399,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 25340
    },
    {
      "epoch": 6.032613223829078,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 25341
    },
    {
      "epoch": 6.032851276557757,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 25342
    },
    {
      "epoch": 6.033089329286437,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1824,
      "step": 25343
    },
    {
      "epoch": 6.033327382015116,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 25344
    },
    {
      "epoch": 6.033565434743796,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 25345
    },
    {
      "epoch": 6.0338034874724755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 25346
    },
    {
      "epoch": 6.034041540201154,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7314,
      "step": 25347
    },
    {
      "epoch": 6.034279592929834,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 25348
    },
    {
      "epoch": 6.0345176456585135,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 25349
    },
    {
      "epoch": 6.034755698387193,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8205,
      "step": 25350
    },
    {
      "epoch": 6.034993751115872,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4047,
      "step": 25351
    },
    {
      "epoch": 6.035231803844551,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 25352
    },
    {
      "epoch": 6.035469856573231,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2461,
      "step": 25353
    },
    {
      "epoch": 6.035707909301911,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 25354
    },
    {
      "epoch": 6.03594596203059,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 25355
    },
    {
      "epoch": 6.036184014759269,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.372,
      "step": 25356
    },
    {
      "epoch": 6.0364220674879485,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 25357
    },
    {
      "epoch": 6.036660120216628,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 25358
    },
    {
      "epoch": 6.036898172945308,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 25359
    },
    {
      "epoch": 6.0371362256739864,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 25360
    },
    {
      "epoch": 6.037374278402666,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 25361
    },
    {
      "epoch": 6.037612331131346,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 25362
    },
    {
      "epoch": 6.037850383860025,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 25363
    },
    {
      "epoch": 6.038088436588704,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2607,
      "step": 25364
    },
    {
      "epoch": 6.038326489317384,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 25365
    },
    {
      "epoch": 6.038564542046063,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 25366
    },
    {
      "epoch": 6.038802594774743,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 25367
    },
    {
      "epoch": 6.039040647503422,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 25368
    },
    {
      "epoch": 6.039278700232101,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 25369
    },
    {
      "epoch": 6.039516752960781,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 25370
    },
    {
      "epoch": 6.03975480568946,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 25371
    },
    {
      "epoch": 6.03999285841814,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 25372
    },
    {
      "epoch": 6.040230911146819,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 25373
    },
    {
      "epoch": 6.040468963875498,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 25374
    },
    {
      "epoch": 6.040707016604178,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 25375
    },
    {
      "epoch": 6.0409450693328575,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 25376
    },
    {
      "epoch": 6.041183122061537,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 25377
    },
    {
      "epoch": 6.041421174790216,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 25378
    },
    {
      "epoch": 6.041659227518895,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 25379
    },
    {
      "epoch": 6.041897280247575,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 25380
    },
    {
      "epoch": 6.042135332976255,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 25381
    },
    {
      "epoch": 6.042373385704933,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 25382
    },
    {
      "epoch": 6.042611438433613,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 25383
    },
    {
      "epoch": 6.0428494911622925,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 25384
    },
    {
      "epoch": 6.043087543890972,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 25385
    },
    {
      "epoch": 6.043325596619651,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 25386
    },
    {
      "epoch": 6.0435636493483305,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 25387
    },
    {
      "epoch": 6.04380170207701,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 25388
    },
    {
      "epoch": 6.04403975480569,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 25389
    },
    {
      "epoch": 6.044277807534369,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 25390
    },
    {
      "epoch": 6.044515860263048,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 25391
    },
    {
      "epoch": 6.044753912991728,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 25392
    },
    {
      "epoch": 6.044991965720407,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 25393
    },
    {
      "epoch": 6.045230018449087,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 25394
    },
    {
      "epoch": 6.0454680711777655,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 25395
    },
    {
      "epoch": 6.045706123906445,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 25396
    },
    {
      "epoch": 6.045944176635125,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 25397
    },
    {
      "epoch": 6.046182229363804,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 25398
    },
    {
      "epoch": 6.046420282092483,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 25399
    },
    {
      "epoch": 6.046658334821163,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 25400
    },
    {
      "epoch": 6.046658334821163,
      "eval_runtime": 0.5449,
      "eval_samples_per_second": 1.835,
      "eval_steps_per_second": 1.835,
      "step": 25400
    },
    {
      "epoch": 6.046896387549842,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 25401
    },
    {
      "epoch": 6.047134440278522,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 25402
    },
    {
      "epoch": 6.0473724930072015,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 25403
    },
    {
      "epoch": 6.04761054573588,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 25404
    },
    {
      "epoch": 6.04784859846456,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 25405
    },
    {
      "epoch": 6.048086651193239,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 25406
    },
    {
      "epoch": 6.048324703921919,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 25407
    },
    {
      "epoch": 6.048562756650598,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 25408
    },
    {
      "epoch": 6.048800809379277,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 25409
    },
    {
      "epoch": 6.049038862107957,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 25410
    },
    {
      "epoch": 6.0492769148366365,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 25411
    },
    {
      "epoch": 6.049514967565316,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 25412
    },
    {
      "epoch": 6.049753020293995,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2442,
      "step": 25413
    },
    {
      "epoch": 6.0499910730226745,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 25414
    },
    {
      "epoch": 6.050229125751354,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 25415
    },
    {
      "epoch": 6.050467178480034,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 25416
    },
    {
      "epoch": 6.050705231208712,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 25417
    },
    {
      "epoch": 6.050943283937392,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 25418
    },
    {
      "epoch": 6.051181336666072,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 25419
    },
    {
      "epoch": 6.051419389394751,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 25420
    },
    {
      "epoch": 6.05165744212343,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 25421
    },
    {
      "epoch": 6.0518954948521095,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 25422
    },
    {
      "epoch": 6.052133547580789,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 25423
    },
    {
      "epoch": 6.052371600309469,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2716,
      "step": 25424
    },
    {
      "epoch": 6.052609653038148,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 25425
    },
    {
      "epoch": 6.052847705766827,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 25426
    },
    {
      "epoch": 6.053085758495507,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.196,
      "step": 25427
    },
    {
      "epoch": 6.053323811224186,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 25428
    },
    {
      "epoch": 6.053561863952866,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 25429
    },
    {
      "epoch": 6.053799916681545,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 25430
    },
    {
      "epoch": 6.054037969410224,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 25431
    },
    {
      "epoch": 6.054276022138904,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 25432
    },
    {
      "epoch": 6.054514074867583,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 25433
    },
    {
      "epoch": 6.054752127596262,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 25434
    },
    {
      "epoch": 6.054990180324942,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3603,
      "step": 25435
    },
    {
      "epoch": 6.055228233053621,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 25436
    },
    {
      "epoch": 6.055466285782301,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 25437
    },
    {
      "epoch": 6.0557043385109806,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 25438
    },
    {
      "epoch": 6.055942391239659,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 25439
    },
    {
      "epoch": 6.056180443968339,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8033,
      "step": 25440
    },
    {
      "epoch": 6.0564184966970185,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 25441
    },
    {
      "epoch": 6.056656549425698,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8078,
      "step": 25442
    },
    {
      "epoch": 6.056894602154377,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 25443
    },
    {
      "epoch": 6.057132654883056,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 25444
    },
    {
      "epoch": 6.057370707611736,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 25445
    },
    {
      "epoch": 6.057608760340416,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 25446
    },
    {
      "epoch": 6.057846813069095,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 25447
    },
    {
      "epoch": 6.058084865797774,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2635,
      "step": 25448
    },
    {
      "epoch": 6.0583229185264535,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 25449
    },
    {
      "epoch": 6.058560971255133,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 25450
    },
    {
      "epoch": 6.058799023983813,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 25451
    },
    {
      "epoch": 6.0590370767124915,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 25452
    },
    {
      "epoch": 6.059275129441171,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 25453
    },
    {
      "epoch": 6.059513182169851,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7767,
      "step": 25454
    },
    {
      "epoch": 6.05975123489853,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 25455
    },
    {
      "epoch": 6.059989287627209,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 25456
    },
    {
      "epoch": 6.060227340355889,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7301,
      "step": 25457
    },
    {
      "epoch": 6.060465393084568,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 25458
    },
    {
      "epoch": 6.060703445813248,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 25459
    },
    {
      "epoch": 6.060941498541927,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3159,
      "step": 25460
    },
    {
      "epoch": 6.061179551270606,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 25461
    },
    {
      "epoch": 6.061417603999286,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 25462
    },
    {
      "epoch": 6.061655656727965,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 25463
    },
    {
      "epoch": 6.061893709456645,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 25464
    },
    {
      "epoch": 6.062131762185324,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 25465
    },
    {
      "epoch": 6.062369814914003,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 25466
    },
    {
      "epoch": 6.062607867642683,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 25467
    },
    {
      "epoch": 6.0628459203713625,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 25468
    },
    {
      "epoch": 6.063083973100042,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 25469
    },
    {
      "epoch": 6.063322025828721,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 25470
    },
    {
      "epoch": 6.0635600785574,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 25471
    },
    {
      "epoch": 6.06379813128608,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 25472
    },
    {
      "epoch": 6.06403618401476,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 25473
    },
    {
      "epoch": 6.064274236743438,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 25474
    },
    {
      "epoch": 6.064512289472118,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3752,
      "step": 25475
    },
    {
      "epoch": 6.064750342200798,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 25476
    },
    {
      "epoch": 6.064988394929477,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 25477
    },
    {
      "epoch": 6.065226447658156,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 25478
    },
    {
      "epoch": 6.0654645003868355,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 25479
    },
    {
      "epoch": 6.065702553115515,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 25480
    },
    {
      "epoch": 6.065940605844195,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 25481
    },
    {
      "epoch": 6.066178658572874,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2188,
      "step": 25482
    },
    {
      "epoch": 6.066416711301553,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 25483
    },
    {
      "epoch": 6.066654764030233,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 25484
    },
    {
      "epoch": 6.066892816758912,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 25485
    },
    {
      "epoch": 6.067130869487592,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 25486
    },
    {
      "epoch": 6.0673689222162706,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 25487
    },
    {
      "epoch": 6.06760697494495,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 25488
    },
    {
      "epoch": 6.06784502767363,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 25489
    },
    {
      "epoch": 6.068083080402309,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9358,
      "step": 25490
    },
    {
      "epoch": 6.068321133130988,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 25491
    },
    {
      "epoch": 6.068559185859668,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 25492
    },
    {
      "epoch": 6.068797238588347,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 25493
    },
    {
      "epoch": 6.069035291317027,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 25494
    },
    {
      "epoch": 6.0692733440457065,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 25495
    },
    {
      "epoch": 6.069511396774385,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3744,
      "step": 25496
    },
    {
      "epoch": 6.069749449503065,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 25497
    },
    {
      "epoch": 6.069987502231744,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8576,
      "step": 25498
    },
    {
      "epoch": 6.070225554960424,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 25499
    },
    {
      "epoch": 6.070463607689103,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 25500
    },
    {
      "epoch": 6.070463607689103,
      "eval_runtime": 0.5232,
      "eval_samples_per_second": 1.911,
      "eval_steps_per_second": 1.911,
      "step": 25500
    },
    {
      "epoch": 6.070701660417782,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 25501
    },
    {
      "epoch": 6.070939713146462,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 25502
    },
    {
      "epoch": 6.071177765875142,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 25503
    },
    {
      "epoch": 6.071415818603821,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 25504
    },
    {
      "epoch": 6.0716538713325,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 25505
    },
    {
      "epoch": 6.0718919240611795,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 25506
    },
    {
      "epoch": 6.072129976789859,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 25507
    },
    {
      "epoch": 6.072368029518539,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.623,
      "step": 25508
    },
    {
      "epoch": 6.072606082247217,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 25509
    },
    {
      "epoch": 6.072844134975897,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3073,
      "step": 25510
    },
    {
      "epoch": 6.073082187704577,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 25511
    },
    {
      "epoch": 6.073320240433256,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 25512
    },
    {
      "epoch": 6.073558293161935,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 25513
    },
    {
      "epoch": 6.073796345890615,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 25514
    },
    {
      "epoch": 6.074034398619294,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2729,
      "step": 25515
    },
    {
      "epoch": 6.074272451347974,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 25516
    },
    {
      "epoch": 6.074510504076653,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 25517
    },
    {
      "epoch": 6.074748556805332,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 25518
    },
    {
      "epoch": 6.074986609534012,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3584,
      "step": 25519
    },
    {
      "epoch": 6.075224662262691,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 25520
    },
    {
      "epoch": 6.075462714991371,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 25521
    },
    {
      "epoch": 6.07570076772005,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 25522
    },
    {
      "epoch": 6.075938820448729,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 25523
    },
    {
      "epoch": 6.076176873177409,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 25524
    },
    {
      "epoch": 6.0764149259060884,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 25525
    },
    {
      "epoch": 6.076652978634767,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 25526
    },
    {
      "epoch": 6.076891031363447,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 25527
    },
    {
      "epoch": 6.077129084092126,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3941,
      "step": 25528
    },
    {
      "epoch": 6.077367136820806,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 25529
    },
    {
      "epoch": 6.077605189549486,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 25530
    },
    {
      "epoch": 6.077843242278164,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 25531
    },
    {
      "epoch": 6.078081295006844,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7009,
      "step": 25532
    },
    {
      "epoch": 6.0783193477355235,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 25533
    },
    {
      "epoch": 6.078557400464203,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 25534
    },
    {
      "epoch": 6.078795453192882,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 25535
    },
    {
      "epoch": 6.079033505921561,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 25536
    },
    {
      "epoch": 6.079271558650241,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 25537
    },
    {
      "epoch": 6.079509611378921,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 25538
    },
    {
      "epoch": 6.0797476641076,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 25539
    },
    {
      "epoch": 6.079985716836279,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 25540
    },
    {
      "epoch": 6.080223769564959,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2439,
      "step": 25541
    },
    {
      "epoch": 6.080461822293638,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 25542
    },
    {
      "epoch": 6.080699875022318,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3985,
      "step": 25543
    },
    {
      "epoch": 6.0809379277509965,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 25544
    },
    {
      "epoch": 6.081175980479676,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 25545
    },
    {
      "epoch": 6.081414033208356,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 25546
    },
    {
      "epoch": 6.081652085937035,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 25547
    },
    {
      "epoch": 6.081890138665714,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 25548
    },
    {
      "epoch": 6.082128191394394,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 25549
    },
    {
      "epoch": 6.082366244123073,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 25550
    },
    {
      "epoch": 6.082604296851753,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 25551
    },
    {
      "epoch": 6.0828423495804325,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 25552
    },
    {
      "epoch": 6.083080402309111,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 25553
    },
    {
      "epoch": 6.083318455037791,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 25554
    },
    {
      "epoch": 6.08355650776647,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 25555
    },
    {
      "epoch": 6.08379456049515,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 25556
    },
    {
      "epoch": 6.084032613223829,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 25557
    },
    {
      "epoch": 6.084270665952508,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 25558
    },
    {
      "epoch": 6.084508718681188,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 25559
    },
    {
      "epoch": 6.0847467714098675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 25560
    },
    {
      "epoch": 6.084984824138546,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 25561
    },
    {
      "epoch": 6.085222876867226,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 25562
    },
    {
      "epoch": 6.0854609295959055,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7662,
      "step": 25563
    },
    {
      "epoch": 6.085698982324585,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7611,
      "step": 25564
    },
    {
      "epoch": 6.085937035053265,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 25565
    },
    {
      "epoch": 6.086175087781943,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 25566
    },
    {
      "epoch": 6.086413140510623,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4091,
      "step": 25567
    },
    {
      "epoch": 6.086651193239303,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 25568
    },
    {
      "epoch": 6.086889245967982,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 25569
    },
    {
      "epoch": 6.087127298696661,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 25570
    },
    {
      "epoch": 6.0873653514253405,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2697,
      "step": 25571
    },
    {
      "epoch": 6.08760340415402,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 25572
    },
    {
      "epoch": 6.0878414568827,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 25573
    },
    {
      "epoch": 6.088079509611379,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 25574
    },
    {
      "epoch": 6.088317562340058,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 25575
    },
    {
      "epoch": 6.088555615068738,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 25576
    },
    {
      "epoch": 6.088793667797417,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 25577
    },
    {
      "epoch": 6.089031720526097,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 25578
    },
    {
      "epoch": 6.089269773254776,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 25579
    },
    {
      "epoch": 6.089507825983455,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 25580
    },
    {
      "epoch": 6.089745878712135,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 25581
    },
    {
      "epoch": 6.089983931440814,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 25582
    },
    {
      "epoch": 6.090221984169493,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 25583
    },
    {
      "epoch": 6.090460036898173,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 25584
    },
    {
      "epoch": 6.090698089626852,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 25585
    },
    {
      "epoch": 6.090936142355532,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 25586
    },
    {
      "epoch": 6.0911741950842115,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 25587
    },
    {
      "epoch": 6.09141224781289,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 25588
    },
    {
      "epoch": 6.09165030054157,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 25589
    },
    {
      "epoch": 6.0918883532702495,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 25590
    },
    {
      "epoch": 6.092126405998929,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 25591
    },
    {
      "epoch": 6.092364458727608,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 25592
    },
    {
      "epoch": 6.092602511456287,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4575,
      "step": 25593
    },
    {
      "epoch": 6.092840564184967,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 25594
    },
    {
      "epoch": 6.093078616913647,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 25595
    },
    {
      "epoch": 6.093316669642326,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 25596
    },
    {
      "epoch": 6.093554722371005,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 25597
    },
    {
      "epoch": 6.0937927750996845,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 25598
    },
    {
      "epoch": 6.094030827828364,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 25599
    },
    {
      "epoch": 6.094268880557044,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 25600
    },
    {
      "epoch": 6.094268880557044,
      "eval_runtime": 0.5301,
      "eval_samples_per_second": 1.886,
      "eval_steps_per_second": 1.886,
      "step": 25600
    },
    {
      "epoch": 6.0945069332857225,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 25601
    },
    {
      "epoch": 6.094744986014402,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 25602
    },
    {
      "epoch": 6.094983038743082,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 25603
    },
    {
      "epoch": 6.095221091471761,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 25604
    },
    {
      "epoch": 6.09545914420044,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 25605
    },
    {
      "epoch": 6.09569719692912,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 25606
    },
    {
      "epoch": 6.095935249657799,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 25607
    },
    {
      "epoch": 6.096173302386479,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3712,
      "step": 25608
    },
    {
      "epoch": 6.096411355115158,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7917,
      "step": 25609
    },
    {
      "epoch": 6.096649407843837,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 25610
    },
    {
      "epoch": 6.096887460572517,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 25611
    },
    {
      "epoch": 6.097125513301196,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 25612
    },
    {
      "epoch": 6.097363566029876,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 25613
    },
    {
      "epoch": 6.097601618758555,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 25614
    },
    {
      "epoch": 6.097839671487234,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 25615
    },
    {
      "epoch": 6.098077724215914,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1572,
      "step": 25616
    },
    {
      "epoch": 6.0983157769445935,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 25617
    },
    {
      "epoch": 6.098553829673272,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 25618
    },
    {
      "epoch": 6.098791882401952,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 25619
    },
    {
      "epoch": 6.099029935130631,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 25620
    },
    {
      "epoch": 6.099267987859311,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 25621
    },
    {
      "epoch": 6.099506040587991,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 25622
    },
    {
      "epoch": 6.099744093316669,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.078,
      "step": 25623
    },
    {
      "epoch": 6.099982146045349,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 25624
    },
    {
      "epoch": 6.1002201987740285,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 25625
    },
    {
      "epoch": 6.100458251502708,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 25626
    },
    {
      "epoch": 6.100696304231387,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 25627
    },
    {
      "epoch": 6.1009343569600665,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 25628
    },
    {
      "epoch": 6.101172409688746,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 25629
    },
    {
      "epoch": 6.101410462417426,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 25630
    },
    {
      "epoch": 6.101648515146105,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 25631
    },
    {
      "epoch": 6.101886567874784,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 25632
    },
    {
      "epoch": 6.102124620603464,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 25633
    },
    {
      "epoch": 6.102362673332143,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 25634
    },
    {
      "epoch": 6.102600726060823,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 25635
    },
    {
      "epoch": 6.1028387787895015,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.365,
      "step": 25636
    },
    {
      "epoch": 6.103076831518181,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 25637
    },
    {
      "epoch": 6.103314884246861,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2544,
      "step": 25638
    },
    {
      "epoch": 6.10355293697554,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 25639
    },
    {
      "epoch": 6.103790989704219,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 25640
    },
    {
      "epoch": 6.104029042432899,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 25641
    },
    {
      "epoch": 6.104267095161578,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 25642
    },
    {
      "epoch": 6.104505147890258,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 25643
    },
    {
      "epoch": 6.1047432006189375,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 25644
    },
    {
      "epoch": 6.104981253347616,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 25645
    },
    {
      "epoch": 6.105219306076296,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 25646
    },
    {
      "epoch": 6.105457358804975,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 25647
    },
    {
      "epoch": 6.105695411533655,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 25648
    },
    {
      "epoch": 6.105933464262334,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 25649
    },
    {
      "epoch": 6.106171516991013,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3169,
      "step": 25650
    },
    {
      "epoch": 6.106409569719693,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 25651
    },
    {
      "epoch": 6.106647622448373,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 25652
    },
    {
      "epoch": 6.106885675177052,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 25653
    },
    {
      "epoch": 6.107123727905731,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 25654
    },
    {
      "epoch": 6.1073617806344105,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 25655
    },
    {
      "epoch": 6.10759983336309,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 25656
    },
    {
      "epoch": 6.10783788609177,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 25657
    },
    {
      "epoch": 6.108075938820448,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 25658
    },
    {
      "epoch": 6.108313991549128,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 25659
    },
    {
      "epoch": 6.108552044277808,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 25660
    },
    {
      "epoch": 6.108790097006487,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 25661
    },
    {
      "epoch": 6.109028149735166,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 25662
    },
    {
      "epoch": 6.1092662024638456,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 25663
    },
    {
      "epoch": 6.109504255192525,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 25664
    },
    {
      "epoch": 6.109742307921205,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 25665
    },
    {
      "epoch": 6.109980360649884,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 25666
    },
    {
      "epoch": 6.110218413378563,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 25667
    },
    {
      "epoch": 6.110456466107243,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 25668
    },
    {
      "epoch": 6.110694518835922,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 25669
    },
    {
      "epoch": 6.110932571564602,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 25670
    },
    {
      "epoch": 6.111170624293281,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 25671
    },
    {
      "epoch": 6.11140867702196,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 25672
    },
    {
      "epoch": 6.11164672975064,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 25673
    },
    {
      "epoch": 6.111884782479319,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.3197,
      "step": 25674
    },
    {
      "epoch": 6.112122835207998,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 25675
    },
    {
      "epoch": 6.112360887936678,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8897,
      "step": 25676
    },
    {
      "epoch": 6.112598940665357,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 25677
    },
    {
      "epoch": 6.112836993394037,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.5725,
      "step": 25678
    },
    {
      "epoch": 6.113075046122717,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 25679
    },
    {
      "epoch": 6.113313098851395,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 25680
    },
    {
      "epoch": 6.113551151580075,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 25681
    },
    {
      "epoch": 6.1137892043087545,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 25682
    },
    {
      "epoch": 6.114027257037434,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 25683
    },
    {
      "epoch": 6.114265309766113,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 25684
    },
    {
      "epoch": 6.114503362494792,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 25685
    },
    {
      "epoch": 6.114741415223472,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 25686
    },
    {
      "epoch": 6.114979467952152,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 25687
    },
    {
      "epoch": 6.115217520680831,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 25688
    },
    {
      "epoch": 6.11545557340951,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 25689
    },
    {
      "epoch": 6.11569362613819,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 25690
    },
    {
      "epoch": 6.115931678866869,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 25691
    },
    {
      "epoch": 6.116169731595549,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 25692
    },
    {
      "epoch": 6.1164077843242275,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2814,
      "step": 25693
    },
    {
      "epoch": 6.116645837052907,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 25694
    },
    {
      "epoch": 6.116883889781587,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 25695
    },
    {
      "epoch": 6.117121942510266,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 25696
    },
    {
      "epoch": 6.117359995238945,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 25697
    },
    {
      "epoch": 6.117598047967625,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 25698
    },
    {
      "epoch": 6.117836100696304,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 25699
    },
    {
      "epoch": 6.118074153424984,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 25700
    },
    {
      "epoch": 6.118074153424984,
      "eval_runtime": 0.5185,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 25700
    },
    {
      "epoch": 6.1183122061536634,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.349,
      "step": 25701
    },
    {
      "epoch": 6.118550258882342,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.5057,
      "step": 25702
    },
    {
      "epoch": 6.118788311611022,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 25703
    },
    {
      "epoch": 6.119026364339701,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6408,
      "step": 25704
    },
    {
      "epoch": 6.119264417068381,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 25705
    },
    {
      "epoch": 6.11950246979706,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8704,
      "step": 25706
    },
    {
      "epoch": 6.119740522525739,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 25707
    },
    {
      "epoch": 6.119978575254419,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 25708
    },
    {
      "epoch": 6.1202166279830985,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 25709
    },
    {
      "epoch": 6.120454680711777,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 25710
    },
    {
      "epoch": 6.120692733440457,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 25711
    },
    {
      "epoch": 6.120930786169136,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 25712
    },
    {
      "epoch": 6.121168838897816,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 25713
    },
    {
      "epoch": 6.121406891626496,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 25714
    },
    {
      "epoch": 6.121644944355174,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 25715
    },
    {
      "epoch": 6.121882997083854,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 25716
    },
    {
      "epoch": 6.122121049812534,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 25717
    },
    {
      "epoch": 6.122359102541213,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 25718
    },
    {
      "epoch": 6.122597155269892,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 25719
    },
    {
      "epoch": 6.1228352079985715,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 25720
    },
    {
      "epoch": 6.123073260727251,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 25721
    },
    {
      "epoch": 6.123311313455931,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 25722
    },
    {
      "epoch": 6.12354936618461,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 25723
    },
    {
      "epoch": 6.123787418913289,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 25724
    },
    {
      "epoch": 6.124025471641969,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 25725
    },
    {
      "epoch": 6.124263524370648,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 25726
    },
    {
      "epoch": 6.124501577099328,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 25727
    },
    {
      "epoch": 6.124739629828007,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 25728
    },
    {
      "epoch": 6.124977682556686,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 25729
    },
    {
      "epoch": 6.125215735285366,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 25730
    },
    {
      "epoch": 6.125453788014045,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 25731
    },
    {
      "epoch": 6.125691840742724,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 25732
    },
    {
      "epoch": 6.125929893471404,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 25733
    },
    {
      "epoch": 6.126167946200083,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3501,
      "step": 25734
    },
    {
      "epoch": 6.126405998928763,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 25735
    },
    {
      "epoch": 6.1266440516574425,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3318,
      "step": 25736
    },
    {
      "epoch": 6.126882104386121,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 25737
    },
    {
      "epoch": 6.127120157114801,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 25738
    },
    {
      "epoch": 6.1273582098434805,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 25739
    },
    {
      "epoch": 6.12759626257216,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 25740
    },
    {
      "epoch": 6.127834315300839,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 25741
    },
    {
      "epoch": 6.128072368029518,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 25742
    },
    {
      "epoch": 6.128310420758198,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 25743
    },
    {
      "epoch": 6.128548473486878,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 25744
    },
    {
      "epoch": 6.128786526215556,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 25745
    },
    {
      "epoch": 6.129024578944236,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2516,
      "step": 25746
    },
    {
      "epoch": 6.1292626316729155,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 25747
    },
    {
      "epoch": 6.129500684401595,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 25748
    },
    {
      "epoch": 6.129738737130275,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 25749
    },
    {
      "epoch": 6.1299767898589534,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 25750
    },
    {
      "epoch": 6.130214842587633,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 25751
    },
    {
      "epoch": 6.130452895316313,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 25752
    },
    {
      "epoch": 6.130690948044992,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 25753
    },
    {
      "epoch": 6.130929000773671,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 25754
    },
    {
      "epoch": 6.131167053502351,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 25755
    },
    {
      "epoch": 6.13140510623103,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 25756
    },
    {
      "epoch": 6.13164315895971,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 25757
    },
    {
      "epoch": 6.131881211688389,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 25758
    },
    {
      "epoch": 6.132119264417068,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8591,
      "step": 25759
    },
    {
      "epoch": 6.132357317145748,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 25760
    },
    {
      "epoch": 6.132595369874427,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3433,
      "step": 25761
    },
    {
      "epoch": 6.132833422603107,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1368,
      "step": 25762
    },
    {
      "epoch": 6.133071475331786,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 25763
    },
    {
      "epoch": 6.133309528060465,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 25764
    },
    {
      "epoch": 6.133547580789145,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 25765
    },
    {
      "epoch": 6.1337856335178245,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 25766
    },
    {
      "epoch": 6.134023686246503,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 25767
    },
    {
      "epoch": 6.134261738975183,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 25768
    },
    {
      "epoch": 6.134499791703862,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 25769
    },
    {
      "epoch": 6.134737844432542,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2987,
      "step": 25770
    },
    {
      "epoch": 6.134975897161222,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 25771
    },
    {
      "epoch": 6.1352139498899,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.3491,
      "step": 25772
    },
    {
      "epoch": 6.13545200261858,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 25773
    },
    {
      "epoch": 6.1356900553472595,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 25774
    },
    {
      "epoch": 6.135928108075939,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 25775
    },
    {
      "epoch": 6.136166160804618,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 25776
    },
    {
      "epoch": 6.1364042135332975,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 25777
    },
    {
      "epoch": 6.136642266261977,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 25778
    },
    {
      "epoch": 6.136880318990657,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 25779
    },
    {
      "epoch": 6.137118371719335,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 25780
    },
    {
      "epoch": 6.137356424448015,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 25781
    },
    {
      "epoch": 6.137594477176695,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 25782
    },
    {
      "epoch": 6.137832529905374,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 25783
    },
    {
      "epoch": 6.138070582634054,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4283,
      "step": 25784
    },
    {
      "epoch": 6.1383086353627325,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3558,
      "step": 25785
    },
    {
      "epoch": 6.138546688091412,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 25786
    },
    {
      "epoch": 6.138784740820092,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 25787
    },
    {
      "epoch": 6.139022793548771,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 25788
    },
    {
      "epoch": 6.13926084627745,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 25789
    },
    {
      "epoch": 6.13949889900613,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 25790
    },
    {
      "epoch": 6.139736951734809,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3681,
      "step": 25791
    },
    {
      "epoch": 6.139975004463489,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2601,
      "step": 25792
    },
    {
      "epoch": 6.1402130571921685,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 25793
    },
    {
      "epoch": 6.140451109920847,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3671,
      "step": 25794
    },
    {
      "epoch": 6.140689162649527,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 25795
    },
    {
      "epoch": 6.140927215378206,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 25796
    },
    {
      "epoch": 6.141165268106886,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 25797
    },
    {
      "epoch": 6.141403320835565,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 25798
    },
    {
      "epoch": 6.141641373564244,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 25799
    },
    {
      "epoch": 6.141879426292924,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 25800
    },
    {
      "epoch": 6.141879426292924,
      "eval_runtime": 0.5274,
      "eval_samples_per_second": 1.896,
      "eval_steps_per_second": 1.896,
      "step": 25800
    },
    {
      "epoch": 6.1421174790216035,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 25801
    },
    {
      "epoch": 6.142355531750282,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 25802
    },
    {
      "epoch": 6.142593584478962,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 25803
    },
    {
      "epoch": 6.1428316372076415,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 25804
    },
    {
      "epoch": 6.143069689936321,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 25805
    },
    {
      "epoch": 6.143307742665001,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 25806
    },
    {
      "epoch": 6.143545795393679,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 25807
    },
    {
      "epoch": 6.143783848122359,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 25808
    },
    {
      "epoch": 6.144021900851039,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 25809
    },
    {
      "epoch": 6.144259953579718,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 25810
    },
    {
      "epoch": 6.144498006308397,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 25811
    },
    {
      "epoch": 6.1447360590370765,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 25812
    },
    {
      "epoch": 6.144974111765756,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 25813
    },
    {
      "epoch": 6.145212164494436,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 25814
    },
    {
      "epoch": 6.1454502172231145,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7382,
      "step": 25815
    },
    {
      "epoch": 6.145688269951794,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 25816
    },
    {
      "epoch": 6.145926322680474,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 25817
    },
    {
      "epoch": 6.146164375409153,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 25818
    },
    {
      "epoch": 6.146402428137833,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 25819
    },
    {
      "epoch": 6.146640480866512,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 25820
    },
    {
      "epoch": 6.146878533595191,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 25821
    },
    {
      "epoch": 6.147116586323871,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 25822
    },
    {
      "epoch": 6.14735463905255,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 25823
    },
    {
      "epoch": 6.147592691781229,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9147,
      "step": 25824
    },
    {
      "epoch": 6.147830744509909,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 25825
    },
    {
      "epoch": 6.148068797238588,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 25826
    },
    {
      "epoch": 6.148306849967268,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 25827
    },
    {
      "epoch": 6.148544902695948,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 25828
    },
    {
      "epoch": 6.148782955424626,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 25829
    },
    {
      "epoch": 6.149021008153306,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 25830
    },
    {
      "epoch": 6.1492590608819855,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 25831
    },
    {
      "epoch": 6.149497113610665,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 25832
    },
    {
      "epoch": 6.149735166339344,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 25833
    },
    {
      "epoch": 6.149973219068023,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 25834
    },
    {
      "epoch": 6.150211271796703,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 25835
    },
    {
      "epoch": 6.150449324525383,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 25836
    },
    {
      "epoch": 6.150687377254062,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 25837
    },
    {
      "epoch": 6.150925429982741,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 25838
    },
    {
      "epoch": 6.1511634827114205,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2496,
      "step": 25839
    },
    {
      "epoch": 6.1514015354401,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 25840
    },
    {
      "epoch": 6.15163958816878,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 25841
    },
    {
      "epoch": 6.1518776408974585,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3059,
      "step": 25842
    },
    {
      "epoch": 6.152115693626138,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.6649,
      "step": 25843
    },
    {
      "epoch": 6.152353746354818,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 25844
    },
    {
      "epoch": 6.152591799083497,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6806,
      "step": 25845
    },
    {
      "epoch": 6.152829851812176,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 25846
    },
    {
      "epoch": 6.153067904540856,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 25847
    },
    {
      "epoch": 6.153305957269535,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3339,
      "step": 25848
    },
    {
      "epoch": 6.153544009998215,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 25849
    },
    {
      "epoch": 6.153782062726894,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 25850
    },
    {
      "epoch": 6.154020115455573,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2588,
      "step": 25851
    },
    {
      "epoch": 6.154258168184253,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 25852
    },
    {
      "epoch": 6.154496220912932,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 25853
    },
    {
      "epoch": 6.154734273641612,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 25854
    },
    {
      "epoch": 6.154972326370291,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7549,
      "step": 25855
    },
    {
      "epoch": 6.15521037909897,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2194,
      "step": 25856
    },
    {
      "epoch": 6.15544843182765,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 25857
    },
    {
      "epoch": 6.1556864845563295,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8503,
      "step": 25858
    },
    {
      "epoch": 6.155924537285008,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 25859
    },
    {
      "epoch": 6.156162590013688,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.3913,
      "step": 25860
    },
    {
      "epoch": 6.156400642742367,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 25861
    },
    {
      "epoch": 6.156638695471047,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 25862
    },
    {
      "epoch": 6.156876748199727,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 25863
    },
    {
      "epoch": 6.157114800928405,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 25864
    },
    {
      "epoch": 6.157352853657085,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 25865
    },
    {
      "epoch": 6.157590906385765,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 25866
    },
    {
      "epoch": 6.157828959114444,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 25867
    },
    {
      "epoch": 6.158067011843123,
      "grad_norm": 7.25,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 25868
    },
    {
      "epoch": 6.1583050645718025,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1664,
      "step": 25869
    },
    {
      "epoch": 6.158543117300482,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 25870
    },
    {
      "epoch": 6.158781170029162,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 25871
    },
    {
      "epoch": 6.159019222757841,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3041,
      "step": 25872
    },
    {
      "epoch": 6.15925727548652,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 25873
    },
    {
      "epoch": 6.1594953282152,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 25874
    },
    {
      "epoch": 6.159733380943879,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 25875
    },
    {
      "epoch": 6.159971433672559,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 25876
    },
    {
      "epoch": 6.1602094864012376,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 25877
    },
    {
      "epoch": 6.160447539129917,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 25878
    },
    {
      "epoch": 6.160685591858597,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 25879
    },
    {
      "epoch": 6.160923644587276,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 25880
    },
    {
      "epoch": 6.161161697315955,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 25881
    },
    {
      "epoch": 6.161399750044635,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 25882
    },
    {
      "epoch": 6.161637802773314,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 25883
    },
    {
      "epoch": 6.161875855501994,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 25884
    },
    {
      "epoch": 6.1621139082306735,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3918,
      "step": 25885
    },
    {
      "epoch": 6.162351960959352,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 25886
    },
    {
      "epoch": 6.162590013688032,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 25887
    },
    {
      "epoch": 6.162828066416711,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 25888
    },
    {
      "epoch": 6.163066119145391,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6532,
      "step": 25889
    },
    {
      "epoch": 6.16330417187407,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 25890
    },
    {
      "epoch": 6.163542224602749,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 25891
    },
    {
      "epoch": 6.163780277331429,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.6824,
      "step": 25892
    },
    {
      "epoch": 6.164018330060109,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 25893
    },
    {
      "epoch": 6.164256382788787,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 25894
    },
    {
      "epoch": 6.164494435517467,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 25895
    },
    {
      "epoch": 6.1647324882461465,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3959,
      "step": 25896
    },
    {
      "epoch": 6.164970540974826,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 25897
    },
    {
      "epoch": 6.165208593703506,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 25898
    },
    {
      "epoch": 6.165446646432184,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 25899
    },
    {
      "epoch": 6.165684699160864,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 25900
    },
    {
      "epoch": 6.165684699160864,
      "eval_runtime": 0.5184,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 25900
    },
    {
      "epoch": 6.165922751889544,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 25901
    },
    {
      "epoch": 6.166160804618223,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 25902
    },
    {
      "epoch": 6.166398857346902,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 25903
    },
    {
      "epoch": 6.166636910075582,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 25904
    },
    {
      "epoch": 6.166874962804261,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 25905
    },
    {
      "epoch": 6.167113015532941,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2779,
      "step": 25906
    },
    {
      "epoch": 6.16735106826162,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.235,
      "step": 25907
    },
    {
      "epoch": 6.167589120990299,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 25908
    },
    {
      "epoch": 6.167827173718979,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7143,
      "step": 25909
    },
    {
      "epoch": 6.168065226447658,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 25910
    },
    {
      "epoch": 6.168303279176338,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 25911
    },
    {
      "epoch": 6.168541331905017,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 25912
    },
    {
      "epoch": 6.168779384633696,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 25913
    },
    {
      "epoch": 6.169017437362376,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 25914
    },
    {
      "epoch": 6.1692554900910554,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 25915
    },
    {
      "epoch": 6.169493542819734,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 25916
    },
    {
      "epoch": 6.169731595548414,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 25917
    },
    {
      "epoch": 6.169969648277093,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2658,
      "step": 25918
    },
    {
      "epoch": 6.170207701005773,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 25919
    },
    {
      "epoch": 6.170445753734453,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 25920
    },
    {
      "epoch": 6.170683806463131,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7303,
      "step": 25921
    },
    {
      "epoch": 6.170921859191811,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 25922
    },
    {
      "epoch": 6.1711599119204905,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 25923
    },
    {
      "epoch": 6.17139796464917,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 25924
    },
    {
      "epoch": 6.171636017377849,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 25925
    },
    {
      "epoch": 6.171874070106528,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 25926
    },
    {
      "epoch": 6.172112122835208,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 25927
    },
    {
      "epoch": 6.172350175563888,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 25928
    },
    {
      "epoch": 6.172588228292566,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 25929
    },
    {
      "epoch": 6.172826281021246,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2929,
      "step": 25930
    },
    {
      "epoch": 6.173064333749926,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 25931
    },
    {
      "epoch": 6.173302386478605,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 25932
    },
    {
      "epoch": 6.173540439207285,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 25933
    },
    {
      "epoch": 6.1737784919359635,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 25934
    },
    {
      "epoch": 6.174016544664643,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 25935
    },
    {
      "epoch": 6.174254597393323,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 25936
    },
    {
      "epoch": 6.174492650122002,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 25937
    },
    {
      "epoch": 6.174730702850681,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 25938
    },
    {
      "epoch": 6.174968755579361,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 25939
    },
    {
      "epoch": 6.17520680830804,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 25940
    },
    {
      "epoch": 6.17544486103672,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 25941
    },
    {
      "epoch": 6.1756829137653995,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 25942
    },
    {
      "epoch": 6.175920966494078,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 25943
    },
    {
      "epoch": 6.176159019222758,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 25944
    },
    {
      "epoch": 6.176397071951437,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 25945
    },
    {
      "epoch": 6.176635124680117,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 25946
    },
    {
      "epoch": 6.176873177408796,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 25947
    },
    {
      "epoch": 6.177111230137475,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 25948
    },
    {
      "epoch": 6.177349282866155,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 25949
    },
    {
      "epoch": 6.1775873355948345,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 25950
    },
    {
      "epoch": 6.177825388323513,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 25951
    },
    {
      "epoch": 6.178063441052193,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 25952
    },
    {
      "epoch": 6.1783014937808725,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8691,
      "step": 25953
    },
    {
      "epoch": 6.178539546509552,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 25954
    },
    {
      "epoch": 6.178777599238232,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 25955
    },
    {
      "epoch": 6.17901565196691,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 25956
    },
    {
      "epoch": 6.17925370469559,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 25957
    },
    {
      "epoch": 6.17949175742427,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 25958
    },
    {
      "epoch": 6.179729810152949,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 25959
    },
    {
      "epoch": 6.179967862881628,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 25960
    },
    {
      "epoch": 6.1802059156103075,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 25961
    },
    {
      "epoch": 6.180443968338987,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 25962
    },
    {
      "epoch": 6.180682021067667,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 25963
    },
    {
      "epoch": 6.1809200737963454,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 25964
    },
    {
      "epoch": 6.181158126525025,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 25965
    },
    {
      "epoch": 6.181396179253705,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 25966
    },
    {
      "epoch": 6.181634231982384,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 25967
    },
    {
      "epoch": 6.181872284711064,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 25968
    },
    {
      "epoch": 6.182110337439743,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 25969
    },
    {
      "epoch": 6.182348390168422,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 25970
    },
    {
      "epoch": 6.182586442897102,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8584,
      "step": 25971
    },
    {
      "epoch": 6.182824495625781,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 25972
    },
    {
      "epoch": 6.18306254835446,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 25973
    },
    {
      "epoch": 6.18330060108314,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 25974
    },
    {
      "epoch": 6.183538653811819,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 25975
    },
    {
      "epoch": 6.183776706540499,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 25976
    },
    {
      "epoch": 6.1840147592691785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 25977
    },
    {
      "epoch": 6.184252811997857,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 25978
    },
    {
      "epoch": 6.184490864726537,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 25979
    },
    {
      "epoch": 6.1847289174552165,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7039,
      "step": 25980
    },
    {
      "epoch": 6.184966970183896,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 25981
    },
    {
      "epoch": 6.185205022912575,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5655,
      "step": 25982
    },
    {
      "epoch": 6.185443075641254,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 25983
    },
    {
      "epoch": 6.185681128369934,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 25984
    },
    {
      "epoch": 6.185919181098614,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 25985
    },
    {
      "epoch": 6.186157233827292,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 25986
    },
    {
      "epoch": 6.186395286555972,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 25987
    },
    {
      "epoch": 6.1866333392846515,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 25988
    },
    {
      "epoch": 6.186871392013331,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 25989
    },
    {
      "epoch": 6.187109444742011,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 25990
    },
    {
      "epoch": 6.1873474974706895,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 25991
    },
    {
      "epoch": 6.187585550199369,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 25992
    },
    {
      "epoch": 6.187823602928049,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 25993
    },
    {
      "epoch": 6.188061655656728,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 25994
    },
    {
      "epoch": 6.188299708385407,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 25995
    },
    {
      "epoch": 6.188537761114087,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 25996
    },
    {
      "epoch": 6.188775813842766,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 25997
    },
    {
      "epoch": 6.189013866571446,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 25998
    },
    {
      "epoch": 6.1892519193001245,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 25999
    },
    {
      "epoch": 6.189489972028804,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 26000
    },
    {
      "epoch": 6.189489972028804,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 26000
    },
    {
      "epoch": 6.189728024757484,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 26001
    },
    {
      "epoch": 6.189966077486163,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 26002
    },
    {
      "epoch": 6.190204130214843,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 26003
    },
    {
      "epoch": 6.190442182943522,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.283,
      "step": 26004
    },
    {
      "epoch": 6.190680235672201,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 26005
    },
    {
      "epoch": 6.190918288400881,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 26006
    },
    {
      "epoch": 6.1911563411295605,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 26007
    },
    {
      "epoch": 6.191394393858239,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 26008
    },
    {
      "epoch": 6.191632446586919,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 26009
    },
    {
      "epoch": 6.191870499315598,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 26010
    },
    {
      "epoch": 6.192108552044278,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 26011
    },
    {
      "epoch": 6.192346604772958,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 26012
    },
    {
      "epoch": 6.192584657501636,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2657,
      "step": 26013
    },
    {
      "epoch": 6.192822710230316,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 26014
    },
    {
      "epoch": 6.1930607629589955,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 26015
    },
    {
      "epoch": 6.193298815687675,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 26016
    },
    {
      "epoch": 6.193536868416354,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 26017
    },
    {
      "epoch": 6.1937749211450335,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 26018
    },
    {
      "epoch": 6.194012973873713,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0145,
      "step": 26019
    },
    {
      "epoch": 6.194251026602393,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1725,
      "step": 26020
    },
    {
      "epoch": 6.194489079331072,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 26021
    },
    {
      "epoch": 6.194727132059751,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.275,
      "step": 26022
    },
    {
      "epoch": 6.194965184788431,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 26023
    },
    {
      "epoch": 6.19520323751711,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 26024
    },
    {
      "epoch": 6.19544129024579,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 26025
    },
    {
      "epoch": 6.1956793429744685,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 26026
    },
    {
      "epoch": 6.195917395703148,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 26027
    },
    {
      "epoch": 6.196155448431828,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 26028
    },
    {
      "epoch": 6.196393501160507,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 26029
    },
    {
      "epoch": 6.196631553889186,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 26030
    },
    {
      "epoch": 6.196869606617866,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 26031
    },
    {
      "epoch": 6.197107659346545,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 26032
    },
    {
      "epoch": 6.197345712075225,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 26033
    },
    {
      "epoch": 6.1975837648039045,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 26034
    },
    {
      "epoch": 6.197821817532583,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.375,
      "step": 26035
    },
    {
      "epoch": 6.198059870261263,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 26036
    },
    {
      "epoch": 6.198297922989942,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6658,
      "step": 26037
    },
    {
      "epoch": 6.198535975718622,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 26038
    },
    {
      "epoch": 6.198774028447301,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 26039
    },
    {
      "epoch": 6.19901208117598,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 26040
    },
    {
      "epoch": 6.19925013390466,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 26041
    },
    {
      "epoch": 6.19948818663334,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 26042
    },
    {
      "epoch": 6.199726239362018,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 26043
    },
    {
      "epoch": 6.199964292090698,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3521,
      "step": 26044
    },
    {
      "epoch": 6.2002023448193775,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.3393,
      "step": 26045
    },
    {
      "epoch": 6.200440397548057,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 26046
    },
    {
      "epoch": 6.200678450276737,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 26047
    },
    {
      "epoch": 6.200916503005415,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 26048
    },
    {
      "epoch": 6.201154555734095,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 26049
    },
    {
      "epoch": 6.201392608462775,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 26050
    },
    {
      "epoch": 6.201630661191454,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2914,
      "step": 26051
    },
    {
      "epoch": 6.201868713920133,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 26052
    },
    {
      "epoch": 6.2021067666488126,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 26053
    },
    {
      "epoch": 6.202344819377492,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 26054
    },
    {
      "epoch": 6.202582872106172,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 26055
    },
    {
      "epoch": 6.202820924834851,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 26056
    },
    {
      "epoch": 6.20305897756353,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 26057
    },
    {
      "epoch": 6.20329703029221,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3333,
      "step": 26058
    },
    {
      "epoch": 6.203535083020889,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 26059
    },
    {
      "epoch": 6.203773135749569,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 26060
    },
    {
      "epoch": 6.204011188478248,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 26061
    },
    {
      "epoch": 6.204249241206927,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2844,
      "step": 26062
    },
    {
      "epoch": 6.204487293935607,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 26063
    },
    {
      "epoch": 6.204725346664286,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 26064
    },
    {
      "epoch": 6.204963399392965,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 26065
    },
    {
      "epoch": 6.205201452121645,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 26066
    },
    {
      "epoch": 6.205439504850324,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.333,
      "step": 26067
    },
    {
      "epoch": 6.205677557579004,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 26068
    },
    {
      "epoch": 6.205915610307684,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 26069
    },
    {
      "epoch": 6.206153663036362,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 26070
    },
    {
      "epoch": 6.206391715765042,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 26071
    },
    {
      "epoch": 6.2066297684937215,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2097,
      "step": 26072
    },
    {
      "epoch": 6.206867821222401,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 26073
    },
    {
      "epoch": 6.20710587395108,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 26074
    },
    {
      "epoch": 6.207343926679759,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 26075
    },
    {
      "epoch": 6.207581979408439,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 26076
    },
    {
      "epoch": 6.207820032137119,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2522,
      "step": 26077
    },
    {
      "epoch": 6.208058084865797,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 26078
    },
    {
      "epoch": 6.208296137594477,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 26079
    },
    {
      "epoch": 6.208534190323157,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 26080
    },
    {
      "epoch": 6.208772243051836,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 26081
    },
    {
      "epoch": 6.209010295780516,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1078,
      "step": 26082
    },
    {
      "epoch": 6.2092483485091945,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 26083
    },
    {
      "epoch": 6.209486401237874,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 26084
    },
    {
      "epoch": 6.209724453966554,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 26085
    },
    {
      "epoch": 6.209962506695233,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 26086
    },
    {
      "epoch": 6.210200559423912,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1782,
      "step": 26087
    },
    {
      "epoch": 6.210438612152592,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 26088
    },
    {
      "epoch": 6.210676664881271,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 26089
    },
    {
      "epoch": 6.210914717609951,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 26090
    },
    {
      "epoch": 6.2111527703386304,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 26091
    },
    {
      "epoch": 6.211390823067309,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9258,
      "step": 26092
    },
    {
      "epoch": 6.211628875795989,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 26093
    },
    {
      "epoch": 6.211866928524668,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3832,
      "step": 26094
    },
    {
      "epoch": 6.212104981253348,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 26095
    },
    {
      "epoch": 6.212343033982027,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 26096
    },
    {
      "epoch": 6.212581086710706,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 26097
    },
    {
      "epoch": 6.212819139439386,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 26098
    },
    {
      "epoch": 6.2130571921680655,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 26099
    },
    {
      "epoch": 6.213295244896744,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 26100
    },
    {
      "epoch": 6.213295244896744,
      "eval_runtime": 0.5155,
      "eval_samples_per_second": 1.94,
      "eval_steps_per_second": 1.94,
      "step": 26100
    },
    {
      "epoch": 6.213533297625424,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 26101
    },
    {
      "epoch": 6.213771350354103,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 26102
    },
    {
      "epoch": 6.214009403082783,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 26103
    },
    {
      "epoch": 6.214247455811463,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 26104
    },
    {
      "epoch": 6.214485508540141,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 26105
    },
    {
      "epoch": 6.214723561268821,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 26106
    },
    {
      "epoch": 6.214961613997501,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 26107
    },
    {
      "epoch": 6.21519966672618,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 26108
    },
    {
      "epoch": 6.215437719454859,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 26109
    },
    {
      "epoch": 6.2156757721835385,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 26110
    },
    {
      "epoch": 6.215913824912218,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 26111
    },
    {
      "epoch": 6.216151877640898,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 26112
    },
    {
      "epoch": 6.216389930369576,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 26113
    },
    {
      "epoch": 6.216627983098256,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 26114
    },
    {
      "epoch": 6.216866035826936,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4093,
      "step": 26115
    },
    {
      "epoch": 6.217104088555615,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 26116
    },
    {
      "epoch": 6.217342141284295,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8166,
      "step": 26117
    },
    {
      "epoch": 6.217580194012974,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 26118
    },
    {
      "epoch": 6.217818246741653,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 26119
    },
    {
      "epoch": 6.218056299470333,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 26120
    },
    {
      "epoch": 6.218294352199012,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 26121
    },
    {
      "epoch": 6.218532404927691,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 26122
    },
    {
      "epoch": 6.218770457656371,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 26123
    },
    {
      "epoch": 6.21900851038505,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8605,
      "step": 26124
    },
    {
      "epoch": 6.21924656311373,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 26125
    },
    {
      "epoch": 6.2194846158424095,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 26126
    },
    {
      "epoch": 6.219722668571088,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 26127
    },
    {
      "epoch": 6.219960721299768,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 26128
    },
    {
      "epoch": 6.2201987740284475,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 26129
    },
    {
      "epoch": 6.220436826757127,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 26130
    },
    {
      "epoch": 6.220674879485806,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 26131
    },
    {
      "epoch": 6.220912932214485,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 26132
    },
    {
      "epoch": 6.221150984943165,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 26133
    },
    {
      "epoch": 6.221389037671845,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 26134
    },
    {
      "epoch": 6.221627090400523,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 26135
    },
    {
      "epoch": 6.221865143129203,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 26136
    },
    {
      "epoch": 6.2221031958578825,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 26137
    },
    {
      "epoch": 6.222341248586562,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 26138
    },
    {
      "epoch": 6.222579301315242,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 26139
    },
    {
      "epoch": 6.2228173540439204,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 26140
    },
    {
      "epoch": 6.2230554067726,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 26141
    },
    {
      "epoch": 6.22329345950128,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 26142
    },
    {
      "epoch": 6.223531512229959,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 26143
    },
    {
      "epoch": 6.223769564958638,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 26144
    },
    {
      "epoch": 6.224007617687318,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 26145
    },
    {
      "epoch": 6.224245670415997,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 26146
    },
    {
      "epoch": 6.224483723144677,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 26147
    },
    {
      "epoch": 6.2247217758733555,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3528,
      "step": 26148
    },
    {
      "epoch": 6.224959828602035,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 26149
    },
    {
      "epoch": 6.225197881330715,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 26150
    },
    {
      "epoch": 6.225435934059394,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 26151
    },
    {
      "epoch": 6.225673986788074,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 26152
    },
    {
      "epoch": 6.225912039516753,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 26153
    },
    {
      "epoch": 6.226150092245432,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 26154
    },
    {
      "epoch": 6.226388144974112,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 26155
    },
    {
      "epoch": 6.2266261977027915,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 26156
    },
    {
      "epoch": 6.22686425043147,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.302,
      "step": 26157
    },
    {
      "epoch": 6.22710230316015,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 26158
    },
    {
      "epoch": 6.227340355888829,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 26159
    },
    {
      "epoch": 6.227578408617509,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 26160
    },
    {
      "epoch": 6.227816461346189,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 26161
    },
    {
      "epoch": 6.228054514074867,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 26162
    },
    {
      "epoch": 6.228292566803547,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 26163
    },
    {
      "epoch": 6.2285306195322265,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8322,
      "step": 26164
    },
    {
      "epoch": 6.228768672260906,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 26165
    },
    {
      "epoch": 6.229006724989585,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 26166
    },
    {
      "epoch": 6.2292447777182645,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2187,
      "step": 26167
    },
    {
      "epoch": 6.229482830446944,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 26168
    },
    {
      "epoch": 6.229720883175624,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 26169
    },
    {
      "epoch": 6.229958935904302,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 26170
    },
    {
      "epoch": 6.230196988632982,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 26171
    },
    {
      "epoch": 6.230435041361662,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1187,
      "step": 26172
    },
    {
      "epoch": 6.230673094090341,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2206,
      "step": 26173
    },
    {
      "epoch": 6.230911146819021,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 26174
    },
    {
      "epoch": 6.2311491995476995,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4,
      "step": 26175
    },
    {
      "epoch": 6.231387252276379,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 26176
    },
    {
      "epoch": 6.231625305005059,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 26177
    },
    {
      "epoch": 6.231863357733738,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 26178
    },
    {
      "epoch": 6.232101410462417,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 26179
    },
    {
      "epoch": 6.232339463191097,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3107,
      "step": 26180
    },
    {
      "epoch": 6.232577515919776,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 26181
    },
    {
      "epoch": 6.232815568648456,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 26182
    },
    {
      "epoch": 6.233053621377135,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 26183
    },
    {
      "epoch": 6.233291674105814,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8813,
      "step": 26184
    },
    {
      "epoch": 6.233529726834494,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 26185
    },
    {
      "epoch": 6.233767779563173,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 26186
    },
    {
      "epoch": 6.234005832291853,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 26187
    },
    {
      "epoch": 6.234243885020532,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8088,
      "step": 26188
    },
    {
      "epoch": 6.234481937749211,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 26189
    },
    {
      "epoch": 6.234719990477891,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 26190
    },
    {
      "epoch": 6.2349580432065705,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 26191
    },
    {
      "epoch": 6.235196095935249,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3669,
      "step": 26192
    },
    {
      "epoch": 6.235434148663929,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 26193
    },
    {
      "epoch": 6.2356722013926085,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 26194
    },
    {
      "epoch": 6.235910254121288,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 26195
    },
    {
      "epoch": 6.236148306849968,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 26196
    },
    {
      "epoch": 6.236386359578646,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 26197
    },
    {
      "epoch": 6.236624412307326,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 26198
    },
    {
      "epoch": 6.236862465036006,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 26199
    },
    {
      "epoch": 6.237100517764685,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 26200
    },
    {
      "epoch": 6.237100517764685,
      "eval_runtime": 0.532,
      "eval_samples_per_second": 1.88,
      "eval_steps_per_second": 1.88,
      "step": 26200
    },
    {
      "epoch": 6.237338570493364,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2926,
      "step": 26201
    },
    {
      "epoch": 6.2375766232220435,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 26202
    },
    {
      "epoch": 6.237814675950723,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 26203
    },
    {
      "epoch": 6.238052728679403,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3825,
      "step": 26204
    },
    {
      "epoch": 6.2382907814080815,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 26205
    },
    {
      "epoch": 6.238528834136761,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 26206
    },
    {
      "epoch": 6.238766886865441,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4831,
      "step": 26207
    },
    {
      "epoch": 6.23900493959412,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8934,
      "step": 26208
    },
    {
      "epoch": 6.2392429923228,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3952,
      "step": 26209
    },
    {
      "epoch": 6.239481045051479,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 26210
    },
    {
      "epoch": 6.239719097780158,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 26211
    },
    {
      "epoch": 6.239957150508838,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 26212
    },
    {
      "epoch": 6.240195203237517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 26213
    },
    {
      "epoch": 6.240433255966196,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 26214
    },
    {
      "epoch": 6.240671308694876,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 26215
    },
    {
      "epoch": 6.240909361423555,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 26216
    },
    {
      "epoch": 6.241147414152235,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.317,
      "step": 26217
    },
    {
      "epoch": 6.241385466880914,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 26218
    },
    {
      "epoch": 6.241623519609593,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 26219
    },
    {
      "epoch": 6.241861572338273,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 26220
    },
    {
      "epoch": 6.2420996250669525,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 26221
    },
    {
      "epoch": 6.242337677795632,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 26222
    },
    {
      "epoch": 6.242575730524311,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 26223
    },
    {
      "epoch": 6.24281378325299,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 26224
    },
    {
      "epoch": 6.24305183598167,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 26225
    },
    {
      "epoch": 6.24328988871035,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 26226
    },
    {
      "epoch": 6.243527941439028,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.397,
      "step": 26227
    },
    {
      "epoch": 6.243765994167708,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3953,
      "step": 26228
    },
    {
      "epoch": 6.2440040468963875,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 26229
    },
    {
      "epoch": 6.244242099625067,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 26230
    },
    {
      "epoch": 6.244480152353747,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 26231
    },
    {
      "epoch": 6.2447182050824255,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 26232
    },
    {
      "epoch": 6.244956257811105,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 26233
    },
    {
      "epoch": 6.245194310539785,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 26234
    },
    {
      "epoch": 6.245432363268464,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 26235
    },
    {
      "epoch": 6.245670415997143,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 26236
    },
    {
      "epoch": 6.245908468725823,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8057,
      "step": 26237
    },
    {
      "epoch": 6.246146521454502,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6874,
      "step": 26238
    },
    {
      "epoch": 6.246384574183182,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 26239
    },
    {
      "epoch": 6.246622626911861,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 26240
    },
    {
      "epoch": 6.24686067964054,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2525,
      "step": 26241
    },
    {
      "epoch": 6.24709873236922,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 26242
    },
    {
      "epoch": 6.247336785097899,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 26243
    },
    {
      "epoch": 6.247574837826579,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 26244
    },
    {
      "epoch": 6.247812890555258,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 26245
    },
    {
      "epoch": 6.248050943283937,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 26246
    },
    {
      "epoch": 6.248288996012617,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 26247
    },
    {
      "epoch": 6.2485270487412965,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7386,
      "step": 26248
    },
    {
      "epoch": 6.248765101469975,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 26249
    },
    {
      "epoch": 6.249003154198655,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 26250
    },
    {
      "epoch": 6.249241206927334,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 26251
    },
    {
      "epoch": 6.249479259656014,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 26252
    },
    {
      "epoch": 6.249717312384694,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 26253
    },
    {
      "epoch": 6.249955365113372,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 26254
    },
    {
      "epoch": 6.250193417842052,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 26255
    },
    {
      "epoch": 6.250431470570732,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1805,
      "step": 26256
    },
    {
      "epoch": 6.250669523299411,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 26257
    },
    {
      "epoch": 6.25090757602809,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 26258
    },
    {
      "epoch": 6.2511456287567695,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 26259
    },
    {
      "epoch": 6.251383681485449,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 26260
    },
    {
      "epoch": 6.251621734214129,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 26261
    },
    {
      "epoch": 6.251859786942807,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 26262
    },
    {
      "epoch": 6.252097839671487,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 26263
    },
    {
      "epoch": 6.252335892400167,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 26264
    },
    {
      "epoch": 6.252573945128846,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 26265
    },
    {
      "epoch": 6.252811997857526,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 26266
    },
    {
      "epoch": 6.2530500505862046,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 26267
    },
    {
      "epoch": 6.253288103314884,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 26268
    },
    {
      "epoch": 6.253526156043564,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 26269
    },
    {
      "epoch": 6.253764208772243,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 26270
    },
    {
      "epoch": 6.254002261500922,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 26271
    },
    {
      "epoch": 6.254240314229602,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 26272
    },
    {
      "epoch": 6.254478366958281,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 26273
    },
    {
      "epoch": 6.254716419686961,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2941,
      "step": 26274
    },
    {
      "epoch": 6.2549544724156405,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 26275
    },
    {
      "epoch": 6.255192525144319,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 26276
    },
    {
      "epoch": 6.255430577872999,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 26277
    },
    {
      "epoch": 6.255668630601678,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 26278
    },
    {
      "epoch": 6.255906683330358,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 26279
    },
    {
      "epoch": 6.256144736059037,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 26280
    },
    {
      "epoch": 6.256382788787716,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 26281
    },
    {
      "epoch": 6.256620841516396,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 26282
    },
    {
      "epoch": 6.256858894245076,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 26283
    },
    {
      "epoch": 6.257096946973754,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 26284
    },
    {
      "epoch": 6.257334999702434,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 26285
    },
    {
      "epoch": 6.2575730524311135,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 26286
    },
    {
      "epoch": 6.257811105159793,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9269,
      "step": 26287
    },
    {
      "epoch": 6.258049157888472,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 26288
    },
    {
      "epoch": 6.258287210617151,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 26289
    },
    {
      "epoch": 6.258525263345831,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 26290
    },
    {
      "epoch": 6.258763316074511,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 26291
    },
    {
      "epoch": 6.25900136880319,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0338,
      "step": 26292
    },
    {
      "epoch": 6.259239421531869,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 26293
    },
    {
      "epoch": 6.259477474260549,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 26294
    },
    {
      "epoch": 6.259715526989228,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 26295
    },
    {
      "epoch": 6.259953579717908,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 26296
    },
    {
      "epoch": 6.2601916324465865,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 26297
    },
    {
      "epoch": 6.260429685175266,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 26298
    },
    {
      "epoch": 6.260667737903946,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.266,
      "step": 26299
    },
    {
      "epoch": 6.260905790632625,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 26300
    },
    {
      "epoch": 6.260905790632625,
      "eval_runtime": 0.5276,
      "eval_samples_per_second": 1.895,
      "eval_steps_per_second": 1.895,
      "step": 26300
    },
    {
      "epoch": 6.261143843361305,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 26301
    },
    {
      "epoch": 6.261381896089984,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 26302
    },
    {
      "epoch": 6.261619948818663,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 26303
    },
    {
      "epoch": 6.261858001547343,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 26304
    },
    {
      "epoch": 6.2620960542760224,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 26305
    },
    {
      "epoch": 6.262334107004701,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7663,
      "step": 26306
    },
    {
      "epoch": 6.262572159733381,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2559,
      "step": 26307
    },
    {
      "epoch": 6.26281021246206,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 26308
    },
    {
      "epoch": 6.26304826519074,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 26309
    },
    {
      "epoch": 6.26328631791942,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 26310
    },
    {
      "epoch": 6.263524370648098,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 26311
    },
    {
      "epoch": 6.263762423376778,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 26312
    },
    {
      "epoch": 6.2640004761054575,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 26313
    },
    {
      "epoch": 6.264238528834137,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 26314
    },
    {
      "epoch": 6.264476581562816,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 26315
    },
    {
      "epoch": 6.264714634291495,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 26316
    },
    {
      "epoch": 6.264952687020175,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 26317
    },
    {
      "epoch": 6.265190739748855,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 26318
    },
    {
      "epoch": 6.265428792477533,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 26319
    },
    {
      "epoch": 6.265666845206213,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1522,
      "step": 26320
    },
    {
      "epoch": 6.265904897934893,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 26321
    },
    {
      "epoch": 6.266142950663572,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 26322
    },
    {
      "epoch": 6.266381003392252,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 26323
    },
    {
      "epoch": 6.2666190561209305,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 26324
    },
    {
      "epoch": 6.26685710884961,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 26325
    },
    {
      "epoch": 6.26709516157829,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 26326
    },
    {
      "epoch": 6.267333214306969,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 26327
    },
    {
      "epoch": 6.267571267035648,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 26328
    },
    {
      "epoch": 6.267809319764328,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 26329
    },
    {
      "epoch": 6.268047372493007,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3371,
      "step": 26330
    },
    {
      "epoch": 6.268285425221687,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 26331
    },
    {
      "epoch": 6.268523477950366,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 26332
    },
    {
      "epoch": 6.268761530679045,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 26333
    },
    {
      "epoch": 6.268999583407725,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4834,
      "step": 26334
    },
    {
      "epoch": 6.269237636136404,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 26335
    },
    {
      "epoch": 6.269475688865084,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 26336
    },
    {
      "epoch": 6.269713741593763,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 26337
    },
    {
      "epoch": 6.269951794322442,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 26338
    },
    {
      "epoch": 6.270189847051122,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 26339
    },
    {
      "epoch": 6.2704278997798015,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 26340
    },
    {
      "epoch": 6.27066595250848,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2685,
      "step": 26341
    },
    {
      "epoch": 6.27090400523716,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 26342
    },
    {
      "epoch": 6.2711420579658395,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 26343
    },
    {
      "epoch": 6.271380110694519,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 26344
    },
    {
      "epoch": 6.271618163423199,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 26345
    },
    {
      "epoch": 6.271856216151877,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 26346
    },
    {
      "epoch": 6.272094268880557,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2752,
      "step": 26347
    },
    {
      "epoch": 6.272332321609237,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0022,
      "step": 26348
    },
    {
      "epoch": 6.272570374337916,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 26349
    },
    {
      "epoch": 6.272808427066595,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 26350
    },
    {
      "epoch": 6.2730464797952745,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 26351
    },
    {
      "epoch": 6.273284532523954,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 26352
    },
    {
      "epoch": 6.273522585252634,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 26353
    },
    {
      "epoch": 6.2737606379813124,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 26354
    },
    {
      "epoch": 6.273998690709992,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 26355
    },
    {
      "epoch": 6.274236743438672,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 26356
    },
    {
      "epoch": 6.274474796167351,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 26357
    },
    {
      "epoch": 6.274712848896031,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 26358
    },
    {
      "epoch": 6.27495090162471,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 26359
    },
    {
      "epoch": 6.275188954353389,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3843,
      "step": 26360
    },
    {
      "epoch": 6.275427007082069,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 26361
    },
    {
      "epoch": 6.275665059810748,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 26362
    },
    {
      "epoch": 6.275903112539427,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 26363
    },
    {
      "epoch": 6.276141165268107,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.5436,
      "step": 26364
    },
    {
      "epoch": 6.276379217996786,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 26365
    },
    {
      "epoch": 6.276617270725466,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 26366
    },
    {
      "epoch": 6.276855323454145,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 26367
    },
    {
      "epoch": 6.277093376182824,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 26368
    },
    {
      "epoch": 6.277331428911504,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 26369
    },
    {
      "epoch": 6.2775694816401835,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 26370
    },
    {
      "epoch": 6.277807534368863,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 26371
    },
    {
      "epoch": 6.278045587097542,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 26372
    },
    {
      "epoch": 6.278283639826221,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 26373
    },
    {
      "epoch": 6.278521692554901,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 26374
    },
    {
      "epoch": 6.278759745283581,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 26375
    },
    {
      "epoch": 6.278997798012259,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 26376
    },
    {
      "epoch": 6.279235850740939,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 26377
    },
    {
      "epoch": 6.2794739034696185,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 26378
    },
    {
      "epoch": 6.279711956198298,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 26379
    },
    {
      "epoch": 6.279950008926978,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 26380
    },
    {
      "epoch": 6.2801880616556565,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 26381
    },
    {
      "epoch": 6.280426114384336,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 26382
    },
    {
      "epoch": 6.280664167113016,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 26383
    },
    {
      "epoch": 6.280902219841695,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9014,
      "step": 26384
    },
    {
      "epoch": 6.281140272570374,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 26385
    },
    {
      "epoch": 6.281378325299054,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 26386
    },
    {
      "epoch": 6.281616378027733,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.391,
      "step": 26387
    },
    {
      "epoch": 6.281854430756413,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 26388
    },
    {
      "epoch": 6.282092483485092,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 26389
    },
    {
      "epoch": 6.282330536213771,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 26390
    },
    {
      "epoch": 6.282568588942451,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 26391
    },
    {
      "epoch": 6.28280664167113,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.7322,
      "step": 26392
    },
    {
      "epoch": 6.28304469439981,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 26393
    },
    {
      "epoch": 6.283282747128489,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 26394
    },
    {
      "epoch": 6.283520799857168,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 26395
    },
    {
      "epoch": 6.283758852585848,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 26396
    },
    {
      "epoch": 6.2839969053145275,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 26397
    },
    {
      "epoch": 6.284234958043206,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 26398
    },
    {
      "epoch": 6.284473010771886,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 26399
    },
    {
      "epoch": 6.284711063500565,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 26400
    },
    {
      "epoch": 6.284711063500565,
      "eval_runtime": 0.5332,
      "eval_samples_per_second": 1.875,
      "eval_steps_per_second": 1.875,
      "step": 26400
    },
    {
      "epoch": 6.284949116229245,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 26401
    },
    {
      "epoch": 6.285187168957924,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 26402
    },
    {
      "epoch": 6.285425221686603,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 26403
    },
    {
      "epoch": 6.285663274415283,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 26404
    },
    {
      "epoch": 6.2859013271439625,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2552,
      "step": 26405
    },
    {
      "epoch": 6.286139379872642,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 26406
    },
    {
      "epoch": 6.286377432601321,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 26407
    },
    {
      "epoch": 6.2866154853300005,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 26408
    },
    {
      "epoch": 6.28685353805868,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 26409
    },
    {
      "epoch": 6.28709159078736,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 26410
    },
    {
      "epoch": 6.287329643516038,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 26411
    },
    {
      "epoch": 6.287567696244718,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2067,
      "step": 26412
    },
    {
      "epoch": 6.287805748973398,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 26413
    },
    {
      "epoch": 6.288043801702077,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 26414
    },
    {
      "epoch": 6.288281854430757,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 26415
    },
    {
      "epoch": 6.2885199071594355,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 26416
    },
    {
      "epoch": 6.288757959888115,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2547,
      "step": 26417
    },
    {
      "epoch": 6.288996012616795,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 26418
    },
    {
      "epoch": 6.289234065345474,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.6523,
      "step": 26419
    },
    {
      "epoch": 6.289472118074153,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 26420
    },
    {
      "epoch": 6.289710170802833,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 26421
    },
    {
      "epoch": 6.289948223531512,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 26422
    },
    {
      "epoch": 6.290186276260192,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 26423
    },
    {
      "epoch": 6.2904243289888715,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 26424
    },
    {
      "epoch": 6.29066238171755,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 26425
    },
    {
      "epoch": 6.29090043444623,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 26426
    },
    {
      "epoch": 6.291138487174909,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 26427
    },
    {
      "epoch": 6.291376539903589,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 26428
    },
    {
      "epoch": 6.291614592632268,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 26429
    },
    {
      "epoch": 6.291852645360947,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 26430
    },
    {
      "epoch": 6.292090698089627,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 26431
    },
    {
      "epoch": 6.292328750818307,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 26432
    },
    {
      "epoch": 6.292566803546985,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 26433
    },
    {
      "epoch": 6.292804856275665,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 26434
    },
    {
      "epoch": 6.2930429090043445,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 26435
    },
    {
      "epoch": 6.293280961733024,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 26436
    },
    {
      "epoch": 6.293519014461703,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 26437
    },
    {
      "epoch": 6.293757067190382,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3468,
      "step": 26438
    },
    {
      "epoch": 6.293995119919062,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 26439
    },
    {
      "epoch": 6.294233172647742,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 26440
    },
    {
      "epoch": 6.294471225376421,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 26441
    },
    {
      "epoch": 6.2947092781051,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 26442
    },
    {
      "epoch": 6.2949473308337796,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 26443
    },
    {
      "epoch": 6.295185383562459,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4076,
      "step": 26444
    },
    {
      "epoch": 6.295423436291139,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 26445
    },
    {
      "epoch": 6.2956614890198175,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 26446
    },
    {
      "epoch": 6.295899541748497,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 26447
    },
    {
      "epoch": 6.296137594477177,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 26448
    },
    {
      "epoch": 6.296375647205856,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 26449
    },
    {
      "epoch": 6.296613699934536,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 26450
    },
    {
      "epoch": 6.296851752663215,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8829,
      "step": 26451
    },
    {
      "epoch": 6.297089805391894,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 26452
    },
    {
      "epoch": 6.297327858120574,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 26453
    },
    {
      "epoch": 6.297565910849253,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 26454
    },
    {
      "epoch": 6.297803963577932,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 26455
    },
    {
      "epoch": 6.298042016306612,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 26456
    },
    {
      "epoch": 6.298280069035291,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 26457
    },
    {
      "epoch": 6.298518121763971,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 26458
    },
    {
      "epoch": 6.298756174492651,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 26459
    },
    {
      "epoch": 6.298994227221329,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 26460
    },
    {
      "epoch": 6.299232279950009,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 26461
    },
    {
      "epoch": 6.2994703326786885,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 26462
    },
    {
      "epoch": 6.299708385407368,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 26463
    },
    {
      "epoch": 6.299946438136047,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 26464
    },
    {
      "epoch": 6.300184490864726,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 26465
    },
    {
      "epoch": 6.300422543593406,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 26466
    },
    {
      "epoch": 6.300660596322086,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 26467
    },
    {
      "epoch": 6.300898649050764,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 26468
    },
    {
      "epoch": 6.301136701779444,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 26469
    },
    {
      "epoch": 6.301374754508124,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 26470
    },
    {
      "epoch": 6.301612807236803,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 26471
    },
    {
      "epoch": 6.301850859965482,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9493,
      "step": 26472
    },
    {
      "epoch": 6.3020889126941615,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 26473
    },
    {
      "epoch": 6.302326965422841,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 26474
    },
    {
      "epoch": 6.302565018151521,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 26475
    },
    {
      "epoch": 6.3028030708802,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3772,
      "step": 26476
    },
    {
      "epoch": 6.303041123608879,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 26477
    },
    {
      "epoch": 6.303279176337559,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 26478
    },
    {
      "epoch": 6.303517229066238,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 26479
    },
    {
      "epoch": 6.303755281794918,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 26480
    },
    {
      "epoch": 6.303993334523597,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 26481
    },
    {
      "epoch": 6.304231387252276,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 26482
    },
    {
      "epoch": 6.304469439980956,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 26483
    },
    {
      "epoch": 6.304707492709635,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 26484
    },
    {
      "epoch": 6.304945545438315,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 26485
    },
    {
      "epoch": 6.305183598166994,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 26486
    },
    {
      "epoch": 6.305421650895673,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 26487
    },
    {
      "epoch": 6.305659703624353,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 26488
    },
    {
      "epoch": 6.3058977563530325,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3822,
      "step": 26489
    },
    {
      "epoch": 6.306135809081711,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2136,
      "step": 26490
    },
    {
      "epoch": 6.306373861810391,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 26491
    },
    {
      "epoch": 6.30661191453907,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 26492
    },
    {
      "epoch": 6.30684996726775,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2636,
      "step": 26493
    },
    {
      "epoch": 6.30708801999643,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3966,
      "step": 26494
    },
    {
      "epoch": 6.307326072725108,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 26495
    },
    {
      "epoch": 6.307564125453788,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 26496
    },
    {
      "epoch": 6.307802178182468,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9017,
      "step": 26497
    },
    {
      "epoch": 6.308040230911147,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 26498
    },
    {
      "epoch": 6.308278283639826,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1762,
      "step": 26499
    },
    {
      "epoch": 6.3085163363685055,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1672,
      "step": 26500
    },
    {
      "epoch": 6.3085163363685055,
      "eval_runtime": 0.535,
      "eval_samples_per_second": 1.869,
      "eval_steps_per_second": 1.869,
      "step": 26500
    },
    {
      "epoch": 6.308754389097185,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 26501
    },
    {
      "epoch": 6.308992441825865,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 26502
    },
    {
      "epoch": 6.309230494554543,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 26503
    },
    {
      "epoch": 6.309468547283223,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 26504
    },
    {
      "epoch": 6.309706600011903,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 26505
    },
    {
      "epoch": 6.309944652740582,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 26506
    },
    {
      "epoch": 6.310182705469262,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 26507
    },
    {
      "epoch": 6.310420758197941,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 26508
    },
    {
      "epoch": 6.31065881092662,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 26509
    },
    {
      "epoch": 6.3108968636553,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 26510
    },
    {
      "epoch": 6.311134916383979,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 26511
    },
    {
      "epoch": 6.311372969112658,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 26512
    },
    {
      "epoch": 6.311611021841338,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 26513
    },
    {
      "epoch": 6.311849074570017,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 26514
    },
    {
      "epoch": 6.312087127298697,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2816,
      "step": 26515
    },
    {
      "epoch": 6.312325180027376,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2167,
      "step": 26516
    },
    {
      "epoch": 6.312563232756055,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 26517
    },
    {
      "epoch": 6.312801285484735,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4384,
      "step": 26518
    },
    {
      "epoch": 6.3130393382134145,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 26519
    },
    {
      "epoch": 6.313277390942094,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1866,
      "step": 26520
    },
    {
      "epoch": 6.313515443670773,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.4871,
      "step": 26521
    },
    {
      "epoch": 6.313753496399452,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 26522
    },
    {
      "epoch": 6.313991549128132,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 26523
    },
    {
      "epoch": 6.314229601856812,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 26524
    },
    {
      "epoch": 6.31446765458549,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1492,
      "step": 26525
    },
    {
      "epoch": 6.31470570731417,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 26526
    },
    {
      "epoch": 6.3149437600428495,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 26527
    },
    {
      "epoch": 6.315181812771529,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 26528
    },
    {
      "epoch": 6.315419865500209,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 26529
    },
    {
      "epoch": 6.3156579182288874,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 26530
    },
    {
      "epoch": 6.315895970957567,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 26531
    },
    {
      "epoch": 6.316134023686247,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 26532
    },
    {
      "epoch": 6.316372076414926,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 26533
    },
    {
      "epoch": 6.316610129143605,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 26534
    },
    {
      "epoch": 6.316848181872285,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 26535
    },
    {
      "epoch": 6.317086234600964,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 26536
    },
    {
      "epoch": 6.317324287329644,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 26537
    },
    {
      "epoch": 6.3175623400583225,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 26538
    },
    {
      "epoch": 6.317800392787002,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 26539
    },
    {
      "epoch": 6.318038445515682,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2872,
      "step": 26540
    },
    {
      "epoch": 6.318276498244361,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 26541
    },
    {
      "epoch": 6.318514550973041,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 26542
    },
    {
      "epoch": 6.31875260370172,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 26543
    },
    {
      "epoch": 6.318990656430399,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3004,
      "step": 26544
    },
    {
      "epoch": 6.319228709159079,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.5219,
      "step": 26545
    },
    {
      "epoch": 6.3194667618877585,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 26546
    },
    {
      "epoch": 6.319704814616437,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5491,
      "step": 26547
    },
    {
      "epoch": 6.319942867345117,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 26548
    },
    {
      "epoch": 6.320180920073796,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 26549
    },
    {
      "epoch": 6.320418972802476,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 26550
    },
    {
      "epoch": 6.320657025531155,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9949,
      "step": 26551
    },
    {
      "epoch": 6.320895078259834,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 26552
    },
    {
      "epoch": 6.321133130988514,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 26553
    },
    {
      "epoch": 6.3213711837171935,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 26554
    },
    {
      "epoch": 6.321609236445873,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8176,
      "step": 26555
    },
    {
      "epoch": 6.321847289174552,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 26556
    },
    {
      "epoch": 6.3220853419032315,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 26557
    },
    {
      "epoch": 6.322323394631911,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 26558
    },
    {
      "epoch": 6.322561447360591,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 26559
    },
    {
      "epoch": 6.322799500089269,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2716,
      "step": 26560
    },
    {
      "epoch": 6.323037552817949,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 26561
    },
    {
      "epoch": 6.323275605546629,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 26562
    },
    {
      "epoch": 6.323513658275308,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8167,
      "step": 26563
    },
    {
      "epoch": 6.323751711003988,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 26564
    },
    {
      "epoch": 6.3239897637326665,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 26565
    },
    {
      "epoch": 6.324227816461346,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 26566
    },
    {
      "epoch": 6.324465869190026,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 26567
    },
    {
      "epoch": 6.324703921918705,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 26568
    },
    {
      "epoch": 6.324941974647384,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 26569
    },
    {
      "epoch": 6.325180027376064,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 26570
    },
    {
      "epoch": 6.325418080104743,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.4863,
      "step": 26571
    },
    {
      "epoch": 6.325656132833423,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 26572
    },
    {
      "epoch": 6.325894185562102,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 26573
    },
    {
      "epoch": 6.326132238290781,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 26574
    },
    {
      "epoch": 6.326370291019461,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8903,
      "step": 26575
    },
    {
      "epoch": 6.32660834374814,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 26576
    },
    {
      "epoch": 6.32684639647682,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 26577
    },
    {
      "epoch": 6.327084449205499,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 26578
    },
    {
      "epoch": 6.327322501934178,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 26579
    },
    {
      "epoch": 6.327560554662858,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2794,
      "step": 26580
    },
    {
      "epoch": 6.3277986073915375,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 26581
    },
    {
      "epoch": 6.328036660120216,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 26582
    },
    {
      "epoch": 6.328274712848896,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3306,
      "step": 26583
    },
    {
      "epoch": 6.3285127655775755,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 26584
    },
    {
      "epoch": 6.328750818306255,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 26585
    },
    {
      "epoch": 6.328988871034934,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 26586
    },
    {
      "epoch": 6.329226923763613,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 26587
    },
    {
      "epoch": 6.329464976492293,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2714,
      "step": 26588
    },
    {
      "epoch": 6.329703029220973,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 26589
    },
    {
      "epoch": 6.329941081949652,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2632,
      "step": 26590
    },
    {
      "epoch": 6.330179134678331,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 26591
    },
    {
      "epoch": 6.3304171874070105,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 26592
    },
    {
      "epoch": 6.33065524013569,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 26593
    },
    {
      "epoch": 6.33089329286437,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 26594
    },
    {
      "epoch": 6.3311313455930485,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3225,
      "step": 26595
    },
    {
      "epoch": 6.331369398321728,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 26596
    },
    {
      "epoch": 6.331607451050408,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 26597
    },
    {
      "epoch": 6.331845503779087,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4806,
      "step": 26598
    },
    {
      "epoch": 6.332083556507767,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 26599
    },
    {
      "epoch": 6.332321609236446,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 26600
    },
    {
      "epoch": 6.332321609236446,
      "eval_runtime": 0.5249,
      "eval_samples_per_second": 1.905,
      "eval_steps_per_second": 1.905,
      "step": 26600
    },
    {
      "epoch": 6.332559661965125,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2043,
      "step": 26601
    },
    {
      "epoch": 6.332797714693805,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 26602
    },
    {
      "epoch": 6.333035767422484,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 26603
    },
    {
      "epoch": 6.333273820151163,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 26604
    },
    {
      "epoch": 6.333511872879843,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 26605
    },
    {
      "epoch": 6.333749925608522,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 26606
    },
    {
      "epoch": 6.333987978337202,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 26607
    },
    {
      "epoch": 6.334226031065882,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 26608
    },
    {
      "epoch": 6.33446408379456,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1058,
      "step": 26609
    },
    {
      "epoch": 6.33470213652324,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 26610
    },
    {
      "epoch": 6.3349401892519195,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 26611
    },
    {
      "epoch": 6.335178241980599,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 26612
    },
    {
      "epoch": 6.335416294709278,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 26613
    },
    {
      "epoch": 6.335654347437957,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 26614
    },
    {
      "epoch": 6.335892400166637,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4622,
      "step": 26615
    },
    {
      "epoch": 6.336130452895317,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 26616
    },
    {
      "epoch": 6.336368505623995,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 26617
    },
    {
      "epoch": 6.336606558352675,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 26618
    },
    {
      "epoch": 6.3368446110813546,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 26619
    },
    {
      "epoch": 6.337082663810034,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 26620
    },
    {
      "epoch": 6.337320716538713,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 26621
    },
    {
      "epoch": 6.3375587692673925,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 26622
    },
    {
      "epoch": 6.337796821996072,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 26623
    },
    {
      "epoch": 6.338034874724752,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 26624
    },
    {
      "epoch": 6.338272927453431,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 26625
    },
    {
      "epoch": 6.33851098018211,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 26626
    },
    {
      "epoch": 6.33874903291079,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 26627
    },
    {
      "epoch": 6.338987085639469,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 26628
    },
    {
      "epoch": 6.339225138368149,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 26629
    },
    {
      "epoch": 6.3394631910968275,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 26630
    },
    {
      "epoch": 6.339701243825507,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 26631
    },
    {
      "epoch": 6.339939296554187,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 26632
    },
    {
      "epoch": 6.340177349282866,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 26633
    },
    {
      "epoch": 6.340415402011546,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6669,
      "step": 26634
    },
    {
      "epoch": 6.340653454740225,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 26635
    },
    {
      "epoch": 6.340891507468904,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 26636
    },
    {
      "epoch": 6.341129560197584,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 26637
    },
    {
      "epoch": 6.3413676129262635,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 26638
    },
    {
      "epoch": 6.341605665654942,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 26639
    },
    {
      "epoch": 6.341843718383622,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 26640
    },
    {
      "epoch": 6.342081771112301,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 26641
    },
    {
      "epoch": 6.342319823840981,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 26642
    },
    {
      "epoch": 6.342557876569661,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9549,
      "step": 26643
    },
    {
      "epoch": 6.342795929298339,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 26644
    },
    {
      "epoch": 6.343033982027019,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 26645
    },
    {
      "epoch": 6.343272034755699,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 26646
    },
    {
      "epoch": 6.343510087484378,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 26647
    },
    {
      "epoch": 6.343748140213057,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 26648
    },
    {
      "epoch": 6.3439861929417365,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 26649
    },
    {
      "epoch": 6.344224245670416,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 26650
    },
    {
      "epoch": 6.344462298399096,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 26651
    },
    {
      "epoch": 6.344700351127774,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 26652
    },
    {
      "epoch": 6.344938403856454,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 26653
    },
    {
      "epoch": 6.345176456585134,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.059,
      "step": 26654
    },
    {
      "epoch": 6.345414509313813,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 26655
    },
    {
      "epoch": 6.345652562042492,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 26656
    },
    {
      "epoch": 6.345890614771172,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7299,
      "step": 26657
    },
    {
      "epoch": 6.346128667499851,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 26658
    },
    {
      "epoch": 6.346366720228531,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 26659
    },
    {
      "epoch": 6.34660477295721,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 26660
    },
    {
      "epoch": 6.346842825685889,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 26661
    },
    {
      "epoch": 6.347080878414569,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 26662
    },
    {
      "epoch": 6.347318931143248,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 26663
    },
    {
      "epoch": 6.347556983871928,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 26664
    },
    {
      "epoch": 6.347795036600607,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 26665
    },
    {
      "epoch": 6.348033089329286,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 26666
    },
    {
      "epoch": 6.348271142057966,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 26667
    },
    {
      "epoch": 6.348509194786645,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 26668
    },
    {
      "epoch": 6.348747247515325,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2219,
      "step": 26669
    },
    {
      "epoch": 6.348985300244004,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 26670
    },
    {
      "epoch": 6.349223352972683,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 26671
    },
    {
      "epoch": 6.349461405701363,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 26672
    },
    {
      "epoch": 6.349699458430043,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 26673
    },
    {
      "epoch": 6.349937511158721,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 26674
    },
    {
      "epoch": 6.350175563887401,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 26675
    },
    {
      "epoch": 6.3504136166160805,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8694,
      "step": 26676
    },
    {
      "epoch": 6.35065166934476,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2237,
      "step": 26677
    },
    {
      "epoch": 6.35088972207344,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 26678
    },
    {
      "epoch": 6.351127774802118,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 26679
    },
    {
      "epoch": 6.351365827530798,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 26680
    },
    {
      "epoch": 6.351603880259478,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 26681
    },
    {
      "epoch": 6.351841932988157,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 26682
    },
    {
      "epoch": 6.352079985716836,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 26683
    },
    {
      "epoch": 6.352318038445516,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 26684
    },
    {
      "epoch": 6.352556091174195,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 26685
    },
    {
      "epoch": 6.352794143902875,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 26686
    },
    {
      "epoch": 6.3530321966315535,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 26687
    },
    {
      "epoch": 6.353270249360233,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 26688
    },
    {
      "epoch": 6.353508302088913,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 26689
    },
    {
      "epoch": 6.353746354817592,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 26690
    },
    {
      "epoch": 6.353984407546271,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 26691
    },
    {
      "epoch": 6.354222460274951,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8717,
      "step": 26692
    },
    {
      "epoch": 6.35446051300363,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 26693
    },
    {
      "epoch": 6.35469856573231,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 26694
    },
    {
      "epoch": 6.3549366184609895,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 26695
    },
    {
      "epoch": 6.355174671189668,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 26696
    },
    {
      "epoch": 6.355412723918348,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 26697
    },
    {
      "epoch": 6.355650776647027,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 26698
    },
    {
      "epoch": 6.355888829375707,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 26699
    },
    {
      "epoch": 6.356126882104386,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 26700
    },
    {
      "epoch": 6.356126882104386,
      "eval_runtime": 0.5229,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 1.912,
      "step": 26700
    },
    {
      "epoch": 6.356364934833065,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 26701
    },
    {
      "epoch": 6.356602987561745,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 26702
    },
    {
      "epoch": 6.3568410402904245,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2313,
      "step": 26703
    },
    {
      "epoch": 6.357079093019104,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 26704
    },
    {
      "epoch": 6.357317145747783,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 26705
    },
    {
      "epoch": 6.357555198476462,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 26706
    },
    {
      "epoch": 6.357793251205142,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 26707
    },
    {
      "epoch": 6.358031303933822,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 26708
    },
    {
      "epoch": 6.3582693566625,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 26709
    },
    {
      "epoch": 6.35850740939118,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 26710
    },
    {
      "epoch": 6.35874546211986,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 26711
    },
    {
      "epoch": 6.358983514848539,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 26712
    },
    {
      "epoch": 6.359221567577219,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.6757,
      "step": 26713
    },
    {
      "epoch": 6.3594596203058975,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2529,
      "step": 26714
    },
    {
      "epoch": 6.359697673034577,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 26715
    },
    {
      "epoch": 6.359935725763257,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 26716
    },
    {
      "epoch": 6.360173778491936,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 26717
    },
    {
      "epoch": 6.360411831220615,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 26718
    },
    {
      "epoch": 6.360649883949295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0075,
      "step": 26719
    },
    {
      "epoch": 6.360887936677974,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 26720
    },
    {
      "epoch": 6.361125989406654,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 26721
    },
    {
      "epoch": 6.361364042135333,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 26722
    },
    {
      "epoch": 6.361602094864012,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 26723
    },
    {
      "epoch": 6.361840147592692,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 26724
    },
    {
      "epoch": 6.362078200321371,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.327,
      "step": 26725
    },
    {
      "epoch": 6.362316253050051,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9393,
      "step": 26726
    },
    {
      "epoch": 6.36255430577873,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 26727
    },
    {
      "epoch": 6.362792358507409,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 26728
    },
    {
      "epoch": 6.363030411236089,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 26729
    },
    {
      "epoch": 6.3632684639647685,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 26730
    },
    {
      "epoch": 6.363506516693447,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 26731
    },
    {
      "epoch": 6.363744569422127,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 26732
    },
    {
      "epoch": 6.3639826221508065,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 26733
    },
    {
      "epoch": 6.364220674879486,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 26734
    },
    {
      "epoch": 6.364458727608165,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 26735
    },
    {
      "epoch": 6.364696780336844,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 26736
    },
    {
      "epoch": 6.364934833065524,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 26737
    },
    {
      "epoch": 6.365172885794204,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 26738
    },
    {
      "epoch": 6.365410938522883,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 26739
    },
    {
      "epoch": 6.365648991251562,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.456,
      "step": 26740
    },
    {
      "epoch": 6.3658870439802415,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 26741
    },
    {
      "epoch": 6.366125096708921,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 26742
    },
    {
      "epoch": 6.366363149437601,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 26743
    },
    {
      "epoch": 6.3666012021662794,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 26744
    },
    {
      "epoch": 6.366839254894959,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 26745
    },
    {
      "epoch": 6.367077307623639,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 26746
    },
    {
      "epoch": 6.367315360352318,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 26747
    },
    {
      "epoch": 6.367553413080998,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 26748
    },
    {
      "epoch": 6.367791465809677,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 26749
    },
    {
      "epoch": 6.368029518538356,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 26750
    },
    {
      "epoch": 6.368267571267036,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 26751
    },
    {
      "epoch": 6.368505623995715,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3281,
      "step": 26752
    },
    {
      "epoch": 6.368743676724394,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 26753
    },
    {
      "epoch": 6.368981729453074,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 26754
    },
    {
      "epoch": 6.369219782181753,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 26755
    },
    {
      "epoch": 6.369457834910433,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 26756
    },
    {
      "epoch": 6.369695887639112,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 26757
    },
    {
      "epoch": 6.369933940367791,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 26758
    },
    {
      "epoch": 6.370171993096471,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 26759
    },
    {
      "epoch": 6.3704100458251505,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 26760
    },
    {
      "epoch": 6.37064809855383,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 26761
    },
    {
      "epoch": 6.370886151282509,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.6474,
      "step": 26762
    },
    {
      "epoch": 6.371124204011188,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 26763
    },
    {
      "epoch": 6.371362256739868,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 26764
    },
    {
      "epoch": 6.371600309468548,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 26765
    },
    {
      "epoch": 6.371838362197226,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1065,
      "step": 26766
    },
    {
      "epoch": 6.372076414925906,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 26767
    },
    {
      "epoch": 6.3723144676545855,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 26768
    },
    {
      "epoch": 6.372552520383265,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 26769
    },
    {
      "epoch": 6.372790573111944,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 26770
    },
    {
      "epoch": 6.3730286258406235,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 26771
    },
    {
      "epoch": 6.373266678569303,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 26772
    },
    {
      "epoch": 6.373504731297983,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 26773
    },
    {
      "epoch": 6.373742784026662,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4707,
      "step": 26774
    },
    {
      "epoch": 6.373980836755341,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 26775
    },
    {
      "epoch": 6.374218889484021,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 26776
    },
    {
      "epoch": 6.3744569422127,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 26777
    },
    {
      "epoch": 6.37469499494138,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 26778
    },
    {
      "epoch": 6.3749330476700585,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3182,
      "step": 26779
    },
    {
      "epoch": 6.375171100398738,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2509,
      "step": 26780
    },
    {
      "epoch": 6.375409153127418,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8244,
      "step": 26781
    },
    {
      "epoch": 6.375647205856097,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 26782
    },
    {
      "epoch": 6.375885258584777,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 26783
    },
    {
      "epoch": 6.376123311313456,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 26784
    },
    {
      "epoch": 6.376361364042135,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 26785
    },
    {
      "epoch": 6.376599416770815,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8123,
      "step": 26786
    },
    {
      "epoch": 6.3768374694994945,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.219,
      "step": 26787
    },
    {
      "epoch": 6.377075522228173,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 26788
    },
    {
      "epoch": 6.377313574956853,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 26789
    },
    {
      "epoch": 6.377551627685532,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 26790
    },
    {
      "epoch": 6.377789680414212,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 26791
    },
    {
      "epoch": 6.378027733142892,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 26792
    },
    {
      "epoch": 6.37826578587157,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 26793
    },
    {
      "epoch": 6.37850383860025,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 26794
    },
    {
      "epoch": 6.3787418913289295,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 26795
    },
    {
      "epoch": 6.378979944057609,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 26796
    },
    {
      "epoch": 6.379217996786288,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8175,
      "step": 26797
    },
    {
      "epoch": 6.3794560495149675,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 26798
    },
    {
      "epoch": 6.379694102243647,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 26799
    },
    {
      "epoch": 6.379932154972327,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 26800
    },
    {
      "epoch": 6.379932154972327,
      "eval_runtime": 0.5168,
      "eval_samples_per_second": 1.935,
      "eval_steps_per_second": 1.935,
      "step": 26800
    },
    {
      "epoch": 6.380170207701005,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 26801
    },
    {
      "epoch": 6.380408260429685,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 26802
    },
    {
      "epoch": 6.380646313158365,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 26803
    },
    {
      "epoch": 6.380884365887044,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 26804
    },
    {
      "epoch": 6.381122418615723,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 26805
    },
    {
      "epoch": 6.3813604713444025,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3641,
      "step": 26806
    },
    {
      "epoch": 6.381598524073082,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 26807
    },
    {
      "epoch": 6.381836576801762,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 26808
    },
    {
      "epoch": 6.382074629530441,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 26809
    },
    {
      "epoch": 6.38231268225912,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 26810
    },
    {
      "epoch": 6.3825507349878,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 26811
    },
    {
      "epoch": 6.382788787716479,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 26812
    },
    {
      "epoch": 6.383026840445159,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 26813
    },
    {
      "epoch": 6.383264893173838,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 26814
    },
    {
      "epoch": 6.383502945902517,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 26815
    },
    {
      "epoch": 6.383740998631197,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 26816
    },
    {
      "epoch": 6.383979051359876,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3011,
      "step": 26817
    },
    {
      "epoch": 6.384217104088556,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 26818
    },
    {
      "epoch": 6.384455156817235,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 26819
    },
    {
      "epoch": 6.384693209545914,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 26820
    },
    {
      "epoch": 6.384931262274594,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 26821
    },
    {
      "epoch": 6.385169315003274,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 26822
    },
    {
      "epoch": 6.385407367731952,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 26823
    },
    {
      "epoch": 6.385645420460632,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 26824
    },
    {
      "epoch": 6.3858834731893115,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8837,
      "step": 26825
    },
    {
      "epoch": 6.386121525917991,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 26826
    },
    {
      "epoch": 6.386359578646671,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 26827
    },
    {
      "epoch": 6.386597631375349,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 26828
    },
    {
      "epoch": 6.386835684104029,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 26829
    },
    {
      "epoch": 6.387073736832709,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 26830
    },
    {
      "epoch": 6.387311789561388,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 26831
    },
    {
      "epoch": 6.387549842290067,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 26832
    },
    {
      "epoch": 6.3877878950187466,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0899,
      "step": 26833
    },
    {
      "epoch": 6.388025947747426,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 26834
    },
    {
      "epoch": 6.388264000476106,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 26835
    },
    {
      "epoch": 6.3885020532047845,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 26836
    },
    {
      "epoch": 6.388740105933464,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 26837
    },
    {
      "epoch": 6.388978158662144,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 26838
    },
    {
      "epoch": 6.389216211390823,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1418,
      "step": 26839
    },
    {
      "epoch": 6.389454264119502,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 26840
    },
    {
      "epoch": 6.389692316848182,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2278,
      "step": 26841
    },
    {
      "epoch": 6.389930369576861,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 26842
    },
    {
      "epoch": 6.390168422305541,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 26843
    },
    {
      "epoch": 6.39040647503422,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 26844
    },
    {
      "epoch": 6.390644527762899,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 26845
    },
    {
      "epoch": 6.390882580491579,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 26846
    },
    {
      "epoch": 6.391120633220258,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 26847
    },
    {
      "epoch": 6.391358685948938,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 26848
    },
    {
      "epoch": 6.391596738677617,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 26849
    },
    {
      "epoch": 6.391834791406296,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 26850
    },
    {
      "epoch": 6.392072844134976,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 26851
    },
    {
      "epoch": 6.3923108968636555,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 26852
    },
    {
      "epoch": 6.392548949592335,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 26853
    },
    {
      "epoch": 6.392787002321014,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 26854
    },
    {
      "epoch": 6.393025055049693,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 26855
    },
    {
      "epoch": 6.393263107778373,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 26856
    },
    {
      "epoch": 6.393501160507053,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 26857
    },
    {
      "epoch": 6.393739213235731,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 26858
    },
    {
      "epoch": 6.393977265964411,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 26859
    },
    {
      "epoch": 6.394215318693091,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2855,
      "step": 26860
    },
    {
      "epoch": 6.39445337142177,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 26861
    },
    {
      "epoch": 6.39469142415045,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 26862
    },
    {
      "epoch": 6.3949294768791285,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 26863
    },
    {
      "epoch": 6.395167529607808,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 26864
    },
    {
      "epoch": 6.395405582336488,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 26865
    },
    {
      "epoch": 6.395643635065167,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 26866
    },
    {
      "epoch": 6.395881687793846,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 26867
    },
    {
      "epoch": 6.396119740522526,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 26868
    },
    {
      "epoch": 6.396357793251205,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 26869
    },
    {
      "epoch": 6.396595845979885,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 26870
    },
    {
      "epoch": 6.396833898708564,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 26871
    },
    {
      "epoch": 6.397071951437243,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 26872
    },
    {
      "epoch": 6.397310004165923,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 26873
    },
    {
      "epoch": 6.397548056894602,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 26874
    },
    {
      "epoch": 6.397786109623281,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 26875
    },
    {
      "epoch": 6.398024162351961,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 26876
    },
    {
      "epoch": 6.39826221508064,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 26877
    },
    {
      "epoch": 6.39850026780932,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 26878
    },
    {
      "epoch": 6.3987383205379995,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 26879
    },
    {
      "epoch": 6.398976373266678,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 26880
    },
    {
      "epoch": 6.399214425995358,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3376,
      "step": 26881
    },
    {
      "epoch": 6.399452478724037,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 26882
    },
    {
      "epoch": 6.399690531452717,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 26883
    },
    {
      "epoch": 6.399928584181396,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 26884
    },
    {
      "epoch": 6.400166636910075,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 26885
    },
    {
      "epoch": 6.400404689638755,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 26886
    },
    {
      "epoch": 6.400642742367435,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 26887
    },
    {
      "epoch": 6.400880795096114,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 26888
    },
    {
      "epoch": 6.401118847824793,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2674,
      "step": 26889
    },
    {
      "epoch": 6.4013569005534725,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.5689,
      "step": 26890
    },
    {
      "epoch": 6.401594953282152,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 26891
    },
    {
      "epoch": 6.401833006010832,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 26892
    },
    {
      "epoch": 6.40207105873951,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 26893
    },
    {
      "epoch": 6.40230911146819,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 26894
    },
    {
      "epoch": 6.40254716419687,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 26895
    },
    {
      "epoch": 6.402785216925549,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 26896
    },
    {
      "epoch": 6.403023269654229,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 26897
    },
    {
      "epoch": 6.403261322382908,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 26898
    },
    {
      "epoch": 6.403499375111587,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 26899
    },
    {
      "epoch": 6.403737427840267,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 26900
    },
    {
      "epoch": 6.403737427840267,
      "eval_runtime": 0.5277,
      "eval_samples_per_second": 1.895,
      "eval_steps_per_second": 1.895,
      "step": 26900
    },
    {
      "epoch": 6.403975480568946,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 26901
    },
    {
      "epoch": 6.404213533297625,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 26902
    },
    {
      "epoch": 6.404451586026305,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9438,
      "step": 26903
    },
    {
      "epoch": 6.404689638754984,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 26904
    },
    {
      "epoch": 6.404927691483664,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 26905
    },
    {
      "epoch": 6.405165744212343,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 26906
    },
    {
      "epoch": 6.405403796941022,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8137,
      "step": 26907
    },
    {
      "epoch": 6.405641849669702,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 26908
    },
    {
      "epoch": 6.4058799023983815,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 26909
    },
    {
      "epoch": 6.406117955127061,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2128,
      "step": 26910
    },
    {
      "epoch": 6.40635600785574,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 26911
    },
    {
      "epoch": 6.406594060584419,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 26912
    },
    {
      "epoch": 6.406832113313099,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 26913
    },
    {
      "epoch": 6.407070166041779,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 26914
    },
    {
      "epoch": 6.407308218770457,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4089,
      "step": 26915
    },
    {
      "epoch": 6.407546271499137,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 26916
    },
    {
      "epoch": 6.4077843242278165,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 26917
    },
    {
      "epoch": 6.408022376956496,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7505,
      "step": 26918
    },
    {
      "epoch": 6.408260429685175,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 26919
    },
    {
      "epoch": 6.4084984824138544,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 26920
    },
    {
      "epoch": 6.408736535142534,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8961,
      "step": 26921
    },
    {
      "epoch": 6.408974587871214,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7988,
      "step": 26922
    },
    {
      "epoch": 6.409212640599893,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 26923
    },
    {
      "epoch": 6.409450693328572,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 26924
    },
    {
      "epoch": 6.409688746057252,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 26925
    },
    {
      "epoch": 6.409926798785931,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 26926
    },
    {
      "epoch": 6.410164851514611,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 26927
    },
    {
      "epoch": 6.4104029042432895,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 26928
    },
    {
      "epoch": 6.410640956971969,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3246,
      "step": 26929
    },
    {
      "epoch": 6.410879009700649,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 26930
    },
    {
      "epoch": 6.411117062429328,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 26931
    },
    {
      "epoch": 6.411355115158008,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 26932
    },
    {
      "epoch": 6.411593167886687,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 26933
    },
    {
      "epoch": 6.411831220615366,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 26934
    },
    {
      "epoch": 6.412069273344046,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8917,
      "step": 26935
    },
    {
      "epoch": 6.4123073260727255,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 26936
    },
    {
      "epoch": 6.412545378801404,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 26937
    },
    {
      "epoch": 6.412783431530084,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 26938
    },
    {
      "epoch": 6.413021484258763,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 26939
    },
    {
      "epoch": 6.413259536987443,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 26940
    },
    {
      "epoch": 6.413497589716122,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 26941
    },
    {
      "epoch": 6.413735642444801,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2099,
      "step": 26942
    },
    {
      "epoch": 6.413973695173481,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 26943
    },
    {
      "epoch": 6.4142117479021605,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 26944
    },
    {
      "epoch": 6.41444980063084,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 26945
    },
    {
      "epoch": 6.414687853359519,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 26946
    },
    {
      "epoch": 6.4149259060881985,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 26947
    },
    {
      "epoch": 6.415163958816878,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 26948
    },
    {
      "epoch": 6.415402011545558,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 26949
    },
    {
      "epoch": 6.415640064274236,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 26950
    },
    {
      "epoch": 6.415878117002916,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3899,
      "step": 26951
    },
    {
      "epoch": 6.416116169731596,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 26952
    },
    {
      "epoch": 6.416354222460275,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 26953
    },
    {
      "epoch": 6.416592275188954,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 26954
    },
    {
      "epoch": 6.4168303279176335,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 26955
    },
    {
      "epoch": 6.417068380646313,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 26956
    },
    {
      "epoch": 6.417306433374993,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2852,
      "step": 26957
    },
    {
      "epoch": 6.417544486103672,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 26958
    },
    {
      "epoch": 6.417782538832351,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 26959
    },
    {
      "epoch": 6.418020591561031,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 26960
    },
    {
      "epoch": 6.41825864428971,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 26961
    },
    {
      "epoch": 6.41849669701839,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 26962
    },
    {
      "epoch": 6.418734749747069,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 26963
    },
    {
      "epoch": 6.418972802475748,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 26964
    },
    {
      "epoch": 6.419210855204428,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 26965
    },
    {
      "epoch": 6.419448907933107,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4125,
      "step": 26966
    },
    {
      "epoch": 6.419686960661787,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3376,
      "step": 26967
    },
    {
      "epoch": 6.419925013390466,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 26968
    },
    {
      "epoch": 6.420163066119145,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 26969
    },
    {
      "epoch": 6.420401118847825,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 26970
    },
    {
      "epoch": 6.4206391715765045,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 26971
    },
    {
      "epoch": 6.420877224305183,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 26972
    },
    {
      "epoch": 6.421115277033863,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 26973
    },
    {
      "epoch": 6.4213533297625425,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.6787,
      "step": 26974
    },
    {
      "epoch": 6.421591382491222,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 26975
    },
    {
      "epoch": 6.421829435219901,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 26976
    },
    {
      "epoch": 6.42206748794858,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 26977
    },
    {
      "epoch": 6.42230554067726,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 26978
    },
    {
      "epoch": 6.42254359340594,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4094,
      "step": 26979
    },
    {
      "epoch": 6.422781646134619,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 26980
    },
    {
      "epoch": 6.423019698863298,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 26981
    },
    {
      "epoch": 6.4232577515919775,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.5188,
      "step": 26982
    },
    {
      "epoch": 6.423495804320657,
      "grad_norm": 6.84375,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 26983
    },
    {
      "epoch": 6.423733857049337,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.3995,
      "step": 26984
    },
    {
      "epoch": 6.4239719097780155,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 26985
    },
    {
      "epoch": 6.424209962506695,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2629,
      "step": 26986
    },
    {
      "epoch": 6.424448015235375,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.331,
      "step": 26987
    },
    {
      "epoch": 6.424686067964054,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3143,
      "step": 26988
    },
    {
      "epoch": 6.424924120692733,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 26989
    },
    {
      "epoch": 6.425162173421413,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 26990
    },
    {
      "epoch": 6.425400226150092,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 26991
    },
    {
      "epoch": 6.425638278878772,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 26992
    },
    {
      "epoch": 6.425876331607451,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7078,
      "step": 26993
    },
    {
      "epoch": 6.42611438433613,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 26994
    },
    {
      "epoch": 6.42635243706481,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3453,
      "step": 26995
    },
    {
      "epoch": 6.426590489793489,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 26996
    },
    {
      "epoch": 6.426828542522169,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 26997
    },
    {
      "epoch": 6.427066595250848,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 26998
    },
    {
      "epoch": 6.427304647979527,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 26999
    },
    {
      "epoch": 6.427542700708207,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 27000
    },
    {
      "epoch": 6.427542700708207,
      "eval_runtime": 0.5282,
      "eval_samples_per_second": 1.893,
      "eval_steps_per_second": 1.893,
      "step": 27000
    },
    {
      "epoch": 6.4277807534368865,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 27001
    },
    {
      "epoch": 6.428018806165566,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 27002
    },
    {
      "epoch": 6.428256858894245,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 27003
    },
    {
      "epoch": 6.428494911622924,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 27004
    },
    {
      "epoch": 6.428732964351604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 27005
    },
    {
      "epoch": 6.428971017080284,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 27006
    },
    {
      "epoch": 6.429209069808962,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 27007
    },
    {
      "epoch": 6.429447122537642,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 27008
    },
    {
      "epoch": 6.4296851752663216,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 27009
    },
    {
      "epoch": 6.429923227995001,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 27010
    },
    {
      "epoch": 6.430161280723681,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 27011
    },
    {
      "epoch": 6.4303993334523595,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 27012
    },
    {
      "epoch": 6.430637386181039,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 27013
    },
    {
      "epoch": 6.430875438909719,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 27014
    },
    {
      "epoch": 6.431113491638398,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 27015
    },
    {
      "epoch": 6.431351544367077,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 27016
    },
    {
      "epoch": 6.431589597095757,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 27017
    },
    {
      "epoch": 6.431827649824436,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 27018
    },
    {
      "epoch": 6.432065702553116,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 27019
    },
    {
      "epoch": 6.4323037552817945,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 27020
    },
    {
      "epoch": 6.432541808010474,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2592,
      "step": 27021
    },
    {
      "epoch": 6.432779860739154,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7932,
      "step": 27022
    },
    {
      "epoch": 6.433017913467833,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 27023
    },
    {
      "epoch": 6.433255966196512,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 27024
    },
    {
      "epoch": 6.433494018925192,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 27025
    },
    {
      "epoch": 6.433732071653871,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 27026
    },
    {
      "epoch": 6.433970124382551,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 27027
    },
    {
      "epoch": 6.4342081771112305,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 27028
    },
    {
      "epoch": 6.434446229839909,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 27029
    },
    {
      "epoch": 6.434684282568589,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 27030
    },
    {
      "epoch": 6.434922335297268,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 27031
    },
    {
      "epoch": 6.435160388025948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 27032
    },
    {
      "epoch": 6.435398440754627,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 27033
    },
    {
      "epoch": 6.435636493483306,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2304,
      "step": 27034
    },
    {
      "epoch": 6.435874546211986,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 27035
    },
    {
      "epoch": 6.436112598940666,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 27036
    },
    {
      "epoch": 6.436350651669345,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 27037
    },
    {
      "epoch": 6.436588704398024,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 27038
    },
    {
      "epoch": 6.4368267571267035,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 27039
    },
    {
      "epoch": 6.437064809855383,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 27040
    },
    {
      "epoch": 6.437302862584063,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 27041
    },
    {
      "epoch": 6.437540915312741,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 27042
    },
    {
      "epoch": 6.437778968041421,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 27043
    },
    {
      "epoch": 6.438017020770101,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 27044
    },
    {
      "epoch": 6.43825507349878,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 27045
    },
    {
      "epoch": 6.43849312622746,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 27046
    },
    {
      "epoch": 6.438731178956139,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 27047
    },
    {
      "epoch": 6.438969231684818,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 27048
    },
    {
      "epoch": 6.439207284413498,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 27049
    },
    {
      "epoch": 6.439445337142177,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9563,
      "step": 27050
    },
    {
      "epoch": 6.439683389870856,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 27051
    },
    {
      "epoch": 6.439921442599536,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 27052
    },
    {
      "epoch": 6.440159495328215,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 27053
    },
    {
      "epoch": 6.440397548056895,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 27054
    },
    {
      "epoch": 6.440635600785574,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 27055
    },
    {
      "epoch": 6.440873653514253,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 27056
    },
    {
      "epoch": 6.441111706242933,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 27057
    },
    {
      "epoch": 6.441349758971612,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 27058
    },
    {
      "epoch": 6.441587811700291,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 27059
    },
    {
      "epoch": 6.441825864428971,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 27060
    },
    {
      "epoch": 6.44206391715765,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 27061
    },
    {
      "epoch": 6.44230196988633,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 27062
    },
    {
      "epoch": 6.44254002261501,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 27063
    },
    {
      "epoch": 6.442778075343688,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.784,
      "step": 27064
    },
    {
      "epoch": 6.443016128072368,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1686,
      "step": 27065
    },
    {
      "epoch": 6.4432541808010475,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.3868,
      "step": 27066
    },
    {
      "epoch": 6.443492233529727,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 27067
    },
    {
      "epoch": 6.443730286258406,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 27068
    },
    {
      "epoch": 6.443968338987085,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 27069
    },
    {
      "epoch": 6.444206391715765,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 27070
    },
    {
      "epoch": 6.444444444444445,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 27071
    },
    {
      "epoch": 6.444682497173124,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 27072
    },
    {
      "epoch": 6.444920549901803,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 27073
    },
    {
      "epoch": 6.445158602630483,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 27074
    },
    {
      "epoch": 6.445396655359162,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 27075
    },
    {
      "epoch": 6.445634708087842,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.8333,
      "step": 27076
    },
    {
      "epoch": 6.4458727608165205,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 27077
    },
    {
      "epoch": 6.4461108135452,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 27078
    },
    {
      "epoch": 6.44634886627388,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 27079
    },
    {
      "epoch": 6.446586919002559,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 27080
    },
    {
      "epoch": 6.446824971731239,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 27081
    },
    {
      "epoch": 6.447063024459918,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9964,
      "step": 27082
    },
    {
      "epoch": 6.447301077188597,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1772,
      "step": 27083
    },
    {
      "epoch": 6.447539129917277,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 27084
    },
    {
      "epoch": 6.4477771826459565,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1584,
      "step": 27085
    },
    {
      "epoch": 6.448015235374635,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 27086
    },
    {
      "epoch": 6.448253288103315,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 27087
    },
    {
      "epoch": 6.448491340831994,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 27088
    },
    {
      "epoch": 6.448729393560674,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 27089
    },
    {
      "epoch": 6.448967446289353,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 27090
    },
    {
      "epoch": 6.449205499018032,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 27091
    },
    {
      "epoch": 6.449443551746712,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 27092
    },
    {
      "epoch": 6.4496816044753915,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 27093
    },
    {
      "epoch": 6.44991965720407,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 27094
    },
    {
      "epoch": 6.45015770993275,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7362,
      "step": 27095
    },
    {
      "epoch": 6.450395762661429,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 27096
    },
    {
      "epoch": 6.450633815390109,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 27097
    },
    {
      "epoch": 6.450871868118789,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 27098
    },
    {
      "epoch": 6.451109920847467,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 27099
    },
    {
      "epoch": 6.451347973576147,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7092,
      "step": 27100
    },
    {
      "epoch": 6.451347973576147,
      "eval_runtime": 0.526,
      "eval_samples_per_second": 1.901,
      "eval_steps_per_second": 1.901,
      "step": 27100
    },
    {
      "epoch": 6.451586026304827,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9703,
      "step": 27101
    },
    {
      "epoch": 6.451824079033506,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 27102
    },
    {
      "epoch": 6.452062131762185,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 27103
    },
    {
      "epoch": 6.4523001844908645,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0575,
      "step": 27104
    },
    {
      "epoch": 6.452538237219544,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 27105
    },
    {
      "epoch": 6.452776289948224,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 27106
    },
    {
      "epoch": 6.453014342676903,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 27107
    },
    {
      "epoch": 6.453252395405582,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 27108
    },
    {
      "epoch": 6.453490448134262,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 27109
    },
    {
      "epoch": 6.453728500862941,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 27110
    },
    {
      "epoch": 6.453966553591621,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 27111
    },
    {
      "epoch": 6.4542046063203,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2751,
      "step": 27112
    },
    {
      "epoch": 6.454442659048979,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 27113
    },
    {
      "epoch": 6.454680711777659,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 27114
    },
    {
      "epoch": 6.454918764506338,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 27115
    },
    {
      "epoch": 6.455156817235018,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 27116
    },
    {
      "epoch": 6.455394869963697,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 27117
    },
    {
      "epoch": 6.455632922692376,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 27118
    },
    {
      "epoch": 6.455870975421056,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 27119
    },
    {
      "epoch": 6.4561090281497355,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 27120
    },
    {
      "epoch": 6.456347080878414,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 27121
    },
    {
      "epoch": 6.456585133607094,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 27122
    },
    {
      "epoch": 6.4568231863357735,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 27123
    },
    {
      "epoch": 6.457061239064453,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 27124
    },
    {
      "epoch": 6.457299291793132,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 27125
    },
    {
      "epoch": 6.457537344521811,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 27126
    },
    {
      "epoch": 6.457775397250491,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 27127
    },
    {
      "epoch": 6.458013449979171,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 27128
    },
    {
      "epoch": 6.45825150270785,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3275,
      "step": 27129
    },
    {
      "epoch": 6.458489555436529,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 27130
    },
    {
      "epoch": 6.4587276081652085,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 27131
    },
    {
      "epoch": 6.458965660893888,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 27132
    },
    {
      "epoch": 6.459203713622568,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.4458,
      "step": 27133
    },
    {
      "epoch": 6.4594417663512465,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 27134
    },
    {
      "epoch": 6.459679819079926,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4044,
      "step": 27135
    },
    {
      "epoch": 6.459917871808606,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 27136
    },
    {
      "epoch": 6.460155924537285,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9277,
      "step": 27137
    },
    {
      "epoch": 6.460393977265964,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 27138
    },
    {
      "epoch": 6.460632029994644,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4099,
      "step": 27139
    },
    {
      "epoch": 6.460870082723323,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 27140
    },
    {
      "epoch": 6.461108135452003,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 27141
    },
    {
      "epoch": 6.461346188180682,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 27142
    },
    {
      "epoch": 6.461584240909361,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 27143
    },
    {
      "epoch": 6.461822293638041,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 27144
    },
    {
      "epoch": 6.46206034636672,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 27145
    },
    {
      "epoch": 6.4622983990954,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 27146
    },
    {
      "epoch": 6.462536451824079,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 27147
    },
    {
      "epoch": 6.462774504552758,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 27148
    },
    {
      "epoch": 6.463012557281438,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 27149
    },
    {
      "epoch": 6.4632506100101175,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 27150
    },
    {
      "epoch": 6.463488662738797,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1876,
      "step": 27151
    },
    {
      "epoch": 6.463726715467476,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 27152
    },
    {
      "epoch": 6.463964768196155,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 27153
    },
    {
      "epoch": 6.464202820924835,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 27154
    },
    {
      "epoch": 6.464440873653515,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3272,
      "step": 27155
    },
    {
      "epoch": 6.464678926382193,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 27156
    },
    {
      "epoch": 6.464916979110873,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 27157
    },
    {
      "epoch": 6.4651550318395525,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2696,
      "step": 27158
    },
    {
      "epoch": 6.465393084568232,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 27159
    },
    {
      "epoch": 6.465631137296911,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 27160
    },
    {
      "epoch": 6.4658691900255905,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 27161
    },
    {
      "epoch": 6.46610724275427,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 27162
    },
    {
      "epoch": 6.46634529548295,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 27163
    },
    {
      "epoch": 6.466583348211629,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0053,
      "step": 27164
    },
    {
      "epoch": 6.466821400940308,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 27165
    },
    {
      "epoch": 6.467059453668988,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3384,
      "step": 27166
    },
    {
      "epoch": 6.467297506397667,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 27167
    },
    {
      "epoch": 6.467535559126347,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 27168
    },
    {
      "epoch": 6.4677736118550255,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 27169
    },
    {
      "epoch": 6.468011664583705,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 27170
    },
    {
      "epoch": 6.468249717312385,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 27171
    },
    {
      "epoch": 6.468487770041064,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 27172
    },
    {
      "epoch": 6.468725822769743,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 27173
    },
    {
      "epoch": 6.468963875498423,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 27174
    },
    {
      "epoch": 6.469201928227102,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 27175
    },
    {
      "epoch": 6.469439980955782,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 27176
    },
    {
      "epoch": 6.4696780336844615,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 27177
    },
    {
      "epoch": 6.46991608641314,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 27178
    },
    {
      "epoch": 6.47015413914182,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 27179
    },
    {
      "epoch": 6.470392191870499,
      "grad_norm": 8.0625,
      "learning_rate": 3e-05,
      "loss": 2.2805,
      "step": 27180
    },
    {
      "epoch": 6.470630244599179,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 27181
    },
    {
      "epoch": 6.470868297327858,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 27182
    },
    {
      "epoch": 6.471106350056537,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 27183
    },
    {
      "epoch": 6.471344402785217,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 27184
    },
    {
      "epoch": 6.4715824555138965,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3475,
      "step": 27185
    },
    {
      "epoch": 6.471820508242576,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 27186
    },
    {
      "epoch": 6.472058560971255,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 27187
    },
    {
      "epoch": 6.4722966136999345,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 27188
    },
    {
      "epoch": 6.472534666428614,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 27189
    },
    {
      "epoch": 6.472772719157294,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 27190
    },
    {
      "epoch": 6.473010771885972,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 27191
    },
    {
      "epoch": 6.473248824614652,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 27192
    },
    {
      "epoch": 6.473486877343332,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 27193
    },
    {
      "epoch": 6.473724930072011,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 27194
    },
    {
      "epoch": 6.473962982800691,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4164,
      "step": 27195
    },
    {
      "epoch": 6.4742010355293695,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 27196
    },
    {
      "epoch": 6.474439088258049,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 27197
    },
    {
      "epoch": 6.474677140986729,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 27198
    },
    {
      "epoch": 6.474915193715408,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 27199
    },
    {
      "epoch": 6.475153246444087,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 27200
    },
    {
      "epoch": 6.475153246444087,
      "eval_runtime": 0.5057,
      "eval_samples_per_second": 1.977,
      "eval_steps_per_second": 1.977,
      "step": 27200
    },
    {
      "epoch": 6.475391299172767,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 27201
    },
    {
      "epoch": 6.475629351901446,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 27202
    },
    {
      "epoch": 6.475867404630126,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 27203
    },
    {
      "epoch": 6.476105457358805,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 27204
    },
    {
      "epoch": 6.476343510087484,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 27205
    },
    {
      "epoch": 6.476581562816164,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 27206
    },
    {
      "epoch": 6.476819615544843,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.323,
      "step": 27207
    },
    {
      "epoch": 6.477057668273522,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 27208
    },
    {
      "epoch": 6.477295721002202,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 27209
    },
    {
      "epoch": 6.477533773730881,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 27210
    },
    {
      "epoch": 6.477771826459561,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 27211
    },
    {
      "epoch": 6.478009879188241,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 27212
    },
    {
      "epoch": 6.478247931916919,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 27213
    },
    {
      "epoch": 6.478485984645599,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3606,
      "step": 27214
    },
    {
      "epoch": 6.4787240373742785,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 27215
    },
    {
      "epoch": 6.478962090102958,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 27216
    },
    {
      "epoch": 6.479200142831637,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 27217
    },
    {
      "epoch": 6.479438195560316,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 27218
    },
    {
      "epoch": 6.479676248288996,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 27219
    },
    {
      "epoch": 6.479914301017676,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 27220
    },
    {
      "epoch": 6.480152353746355,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 27221
    },
    {
      "epoch": 6.480390406475034,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 27222
    },
    {
      "epoch": 6.4806284592037136,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 27223
    },
    {
      "epoch": 6.480866511932393,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 27224
    },
    {
      "epoch": 6.481104564661073,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 27225
    },
    {
      "epoch": 6.4813426173897515,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 27226
    },
    {
      "epoch": 6.481580670118431,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 27227
    },
    {
      "epoch": 6.481818722847111,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.6018,
      "step": 27228
    },
    {
      "epoch": 6.48205677557579,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 27229
    },
    {
      "epoch": 6.48229482830447,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 27230
    },
    {
      "epoch": 6.482532881033149,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8725,
      "step": 27231
    },
    {
      "epoch": 6.482770933761828,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.5578,
      "step": 27232
    },
    {
      "epoch": 6.483008986490508,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 27233
    },
    {
      "epoch": 6.483247039219187,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 27234
    },
    {
      "epoch": 6.483485091947866,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 27235
    },
    {
      "epoch": 6.483723144676546,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 27236
    },
    {
      "epoch": 6.483961197405225,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 27237
    },
    {
      "epoch": 6.484199250133905,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 27238
    },
    {
      "epoch": 6.484437302862584,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 27239
    },
    {
      "epoch": 6.484675355591263,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 27240
    },
    {
      "epoch": 6.484913408319943,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 27241
    },
    {
      "epoch": 6.4851514610486225,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 27242
    },
    {
      "epoch": 6.485389513777301,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 27243
    },
    {
      "epoch": 6.485627566505981,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 27244
    },
    {
      "epoch": 6.48586561923466,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 27245
    },
    {
      "epoch": 6.48610367196334,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 27246
    },
    {
      "epoch": 6.48634172469202,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 27247
    },
    {
      "epoch": 6.486579777420698,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 27248
    },
    {
      "epoch": 6.486817830149378,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.7106,
      "step": 27249
    },
    {
      "epoch": 6.487055882878058,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 27250
    },
    {
      "epoch": 6.487293935606737,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 27251
    },
    {
      "epoch": 6.487531988335416,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4743,
      "step": 27252
    },
    {
      "epoch": 6.4877700410640955,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 27253
    },
    {
      "epoch": 6.488008093792775,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 27254
    },
    {
      "epoch": 6.488246146521455,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 27255
    },
    {
      "epoch": 6.488484199250134,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 27256
    },
    {
      "epoch": 6.488722251978813,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 27257
    },
    {
      "epoch": 6.488960304707493,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 27258
    },
    {
      "epoch": 6.489198357436172,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3781,
      "step": 27259
    },
    {
      "epoch": 6.489436410164852,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9733,
      "step": 27260
    },
    {
      "epoch": 6.489674462893531,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 27261
    },
    {
      "epoch": 6.48991251562221,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 27262
    },
    {
      "epoch": 6.49015056835089,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.225,
      "step": 27263
    },
    {
      "epoch": 6.490388621079569,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 27264
    },
    {
      "epoch": 6.490626673808249,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 27265
    },
    {
      "epoch": 6.490864726536928,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 27266
    },
    {
      "epoch": 6.491102779265607,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3989,
      "step": 27267
    },
    {
      "epoch": 6.491340831994287,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8196,
      "step": 27268
    },
    {
      "epoch": 6.4915788847229665,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 27269
    },
    {
      "epoch": 6.491816937451645,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 27270
    },
    {
      "epoch": 6.492054990180325,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 27271
    },
    {
      "epoch": 6.492293042909004,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 27272
    },
    {
      "epoch": 6.492531095637684,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 27273
    },
    {
      "epoch": 6.492769148366363,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 27274
    },
    {
      "epoch": 6.493007201095042,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 27275
    },
    {
      "epoch": 6.493245253823722,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3775,
      "step": 27276
    },
    {
      "epoch": 6.493483306552402,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 27277
    },
    {
      "epoch": 6.49372135928108,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 27278
    },
    {
      "epoch": 6.49395941200976,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 27279
    },
    {
      "epoch": 6.4941974647384395,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 27280
    },
    {
      "epoch": 6.494435517467119,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 27281
    },
    {
      "epoch": 6.494673570195799,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 27282
    },
    {
      "epoch": 6.494911622924477,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 27283
    },
    {
      "epoch": 6.495149675653157,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 27284
    },
    {
      "epoch": 6.495387728381837,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 27285
    },
    {
      "epoch": 6.495625781110516,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.5165,
      "step": 27286
    },
    {
      "epoch": 6.495863833839195,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 27287
    },
    {
      "epoch": 6.496101886567875,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 27288
    },
    {
      "epoch": 6.496339939296554,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 27289
    },
    {
      "epoch": 6.496577992025234,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 27290
    },
    {
      "epoch": 6.496816044753913,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8745,
      "step": 27291
    },
    {
      "epoch": 6.497054097482592,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 27292
    },
    {
      "epoch": 6.497292150211272,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 27293
    },
    {
      "epoch": 6.497530202939951,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 27294
    },
    {
      "epoch": 6.497768255668631,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 27295
    },
    {
      "epoch": 6.49800630839731,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3165,
      "step": 27296
    },
    {
      "epoch": 6.498244361125989,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 27297
    },
    {
      "epoch": 6.498482413854669,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 27298
    },
    {
      "epoch": 6.4987204665833485,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 27299
    },
    {
      "epoch": 6.498958519312028,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2531,
      "step": 27300
    },
    {
      "epoch": 6.498958519312028,
      "eval_runtime": 0.5092,
      "eval_samples_per_second": 1.964,
      "eval_steps_per_second": 1.964,
      "step": 27300
    },
    {
      "epoch": 6.499196572040707,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 27301
    },
    {
      "epoch": 6.499434624769386,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 27302
    },
    {
      "epoch": 6.499672677498066,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 27303
    },
    {
      "epoch": 6.499910730226746,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 27304
    },
    {
      "epoch": 6.500148782955424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 27305
    },
    {
      "epoch": 6.500386835684104,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4066,
      "step": 27306
    },
    {
      "epoch": 6.5006248884127835,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 27307
    },
    {
      "epoch": 6.500862941141463,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 27308
    },
    {
      "epoch": 6.501100993870143,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 27309
    },
    {
      "epoch": 6.5013390465988214,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5078,
      "step": 27310
    },
    {
      "epoch": 6.501577099327501,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3793,
      "step": 27311
    },
    {
      "epoch": 6.501815152056181,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 27312
    },
    {
      "epoch": 6.502053204784859,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 27313
    },
    {
      "epoch": 6.502291257513539,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 27314
    },
    {
      "epoch": 6.502529310242219,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 27315
    },
    {
      "epoch": 6.502767362970898,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 27316
    },
    {
      "epoch": 6.503005415699578,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 27317
    },
    {
      "epoch": 6.5032434684282565,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 27318
    },
    {
      "epoch": 6.503481521156936,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 27319
    },
    {
      "epoch": 6.503719573885616,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 27320
    },
    {
      "epoch": 6.503957626614295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 27321
    },
    {
      "epoch": 6.504195679342974,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 27322
    },
    {
      "epoch": 6.504433732071654,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 27323
    },
    {
      "epoch": 6.504671784800333,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 27324
    },
    {
      "epoch": 6.504909837529013,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 27325
    },
    {
      "epoch": 6.5051478902576925,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 27326
    },
    {
      "epoch": 6.505385942986371,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 27327
    },
    {
      "epoch": 6.505623995715051,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 27328
    },
    {
      "epoch": 6.50586204844373,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 27329
    },
    {
      "epoch": 6.50610010117241,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 27330
    },
    {
      "epoch": 6.506338153901089,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.862,
      "step": 27331
    },
    {
      "epoch": 6.506576206629768,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 27332
    },
    {
      "epoch": 6.506814259358448,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 27333
    },
    {
      "epoch": 6.5070523120871275,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 27334
    },
    {
      "epoch": 6.507290364815807,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 27335
    },
    {
      "epoch": 6.507528417544486,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 27336
    },
    {
      "epoch": 6.5077664702731655,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 27337
    },
    {
      "epoch": 6.508004523001845,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 27338
    },
    {
      "epoch": 6.508242575730525,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3482,
      "step": 27339
    },
    {
      "epoch": 6.508480628459203,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4665,
      "step": 27340
    },
    {
      "epoch": 6.508718681187883,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 27341
    },
    {
      "epoch": 6.508956733916563,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 27342
    },
    {
      "epoch": 6.509194786645242,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9888,
      "step": 27343
    },
    {
      "epoch": 6.509432839373922,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3495,
      "step": 27344
    },
    {
      "epoch": 6.5096708921026005,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 27345
    },
    {
      "epoch": 6.50990894483128,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 27346
    },
    {
      "epoch": 6.51014699755996,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 27347
    },
    {
      "epoch": 6.5103850502886385,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 27348
    },
    {
      "epoch": 6.510623103017318,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 27349
    },
    {
      "epoch": 6.510861155745998,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 27350
    },
    {
      "epoch": 6.511099208474677,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 27351
    },
    {
      "epoch": 6.511337261203357,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 27352
    },
    {
      "epoch": 6.511575313932036,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 27353
    },
    {
      "epoch": 6.511813366660715,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 27354
    },
    {
      "epoch": 6.512051419389395,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 27355
    },
    {
      "epoch": 6.512289472118074,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6235,
      "step": 27356
    },
    {
      "epoch": 6.512527524846753,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 27357
    },
    {
      "epoch": 6.512765577575433,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 27358
    },
    {
      "epoch": 6.513003630304112,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 27359
    },
    {
      "epoch": 6.513241683032792,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 27360
    },
    {
      "epoch": 6.5134797357614715,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.5929,
      "step": 27361
    },
    {
      "epoch": 6.51371778849015,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 27362
    },
    {
      "epoch": 6.51395584121883,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3826,
      "step": 27363
    },
    {
      "epoch": 6.5141938939475095,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 27364
    },
    {
      "epoch": 6.514431946676189,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 27365
    },
    {
      "epoch": 6.514669999404868,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 27366
    },
    {
      "epoch": 6.514908052133547,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1534,
      "step": 27367
    },
    {
      "epoch": 6.515146104862227,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 27368
    },
    {
      "epoch": 6.515384157590907,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 27369
    },
    {
      "epoch": 6.515622210319586,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 27370
    },
    {
      "epoch": 6.515860263048265,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 27371
    },
    {
      "epoch": 6.5160983157769445,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 27372
    },
    {
      "epoch": 6.516336368505624,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 27373
    },
    {
      "epoch": 6.516574421234304,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 27374
    },
    {
      "epoch": 6.5168124739629825,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 27375
    },
    {
      "epoch": 6.517050526691662,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 27376
    },
    {
      "epoch": 6.517288579420342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 27377
    },
    {
      "epoch": 6.517526632149021,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 27378
    },
    {
      "epoch": 6.517764684877701,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 27379
    },
    {
      "epoch": 6.51800273760638,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4191,
      "step": 27380
    },
    {
      "epoch": 6.518240790335059,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 27381
    },
    {
      "epoch": 6.518478843063739,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 27382
    },
    {
      "epoch": 6.5187168957924175,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 27383
    },
    {
      "epoch": 6.518954948521097,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 27384
    },
    {
      "epoch": 6.519193001249777,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 27385
    },
    {
      "epoch": 6.519431053978456,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 27386
    },
    {
      "epoch": 6.519669106707136,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 27387
    },
    {
      "epoch": 6.519907159435815,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 27388
    },
    {
      "epoch": 6.520145212164494,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 27389
    },
    {
      "epoch": 6.520383264893174,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 27390
    },
    {
      "epoch": 6.5206213176218535,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 27391
    },
    {
      "epoch": 6.520859370350532,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 27392
    },
    {
      "epoch": 6.521097423079212,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3823,
      "step": 27393
    },
    {
      "epoch": 6.521335475807891,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 27394
    },
    {
      "epoch": 6.521573528536571,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3354,
      "step": 27395
    },
    {
      "epoch": 6.521811581265251,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 27396
    },
    {
      "epoch": 6.522049633993929,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 27397
    },
    {
      "epoch": 6.522287686722609,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 27398
    },
    {
      "epoch": 6.5225257394512886,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8775,
      "step": 27399
    },
    {
      "epoch": 6.522763792179968,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 27400
    },
    {
      "epoch": 6.522763792179968,
      "eval_runtime": 0.5137,
      "eval_samples_per_second": 1.947,
      "eval_steps_per_second": 1.947,
      "step": 27400
    },
    {
      "epoch": 6.523001844908647,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8396,
      "step": 27401
    },
    {
      "epoch": 6.5232398976373265,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 27402
    },
    {
      "epoch": 6.523477950366006,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 27403
    },
    {
      "epoch": 6.523716003094686,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 27404
    },
    {
      "epoch": 6.523954055823365,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 27405
    },
    {
      "epoch": 6.524192108552044,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 27406
    },
    {
      "epoch": 6.524430161280724,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 27407
    },
    {
      "epoch": 6.524668214009403,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 27408
    },
    {
      "epoch": 6.524906266738083,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 27409
    },
    {
      "epoch": 6.5251443194667615,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 27410
    },
    {
      "epoch": 6.525382372195441,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 27411
    },
    {
      "epoch": 6.525620424924121,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 27412
    },
    {
      "epoch": 6.5258584776528,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.5383,
      "step": 27413
    },
    {
      "epoch": 6.52609653038148,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 27414
    },
    {
      "epoch": 6.526334583110159,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 27415
    },
    {
      "epoch": 6.526572635838838,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 27416
    },
    {
      "epoch": 6.526810688567518,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 27417
    },
    {
      "epoch": 6.527048741296197,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 27418
    },
    {
      "epoch": 6.527286794024876,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 27419
    },
    {
      "epoch": 6.527524846753556,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 27420
    },
    {
      "epoch": 6.527762899482235,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 27421
    },
    {
      "epoch": 6.528000952210915,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 27422
    },
    {
      "epoch": 6.528239004939594,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0021,
      "step": 27423
    },
    {
      "epoch": 6.528477057668273,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3997,
      "step": 27424
    },
    {
      "epoch": 6.528715110396953,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2452,
      "step": 27425
    },
    {
      "epoch": 6.528953163125633,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 27426
    },
    {
      "epoch": 6.529191215854311,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 27427
    },
    {
      "epoch": 6.529429268582991,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 27428
    },
    {
      "epoch": 6.5296673213116705,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 27429
    },
    {
      "epoch": 6.52990537404035,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 27430
    },
    {
      "epoch": 6.53014342676903,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 27431
    },
    {
      "epoch": 6.530381479497708,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 27432
    },
    {
      "epoch": 6.530619532226388,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 27433
    },
    {
      "epoch": 6.530857584955068,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3736,
      "step": 27434
    },
    {
      "epoch": 6.531095637683747,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 27435
    },
    {
      "epoch": 6.531333690412426,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 27436
    },
    {
      "epoch": 6.531571743141106,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 27437
    },
    {
      "epoch": 6.531809795869785,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 27438
    },
    {
      "epoch": 6.532047848598465,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 27439
    },
    {
      "epoch": 6.532285901327144,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 27440
    },
    {
      "epoch": 6.532523954055823,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 27441
    },
    {
      "epoch": 6.532762006784503,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 27442
    },
    {
      "epoch": 6.533000059513182,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 27443
    },
    {
      "epoch": 6.533238112241862,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4321,
      "step": 27444
    },
    {
      "epoch": 6.533476164970541,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 27445
    },
    {
      "epoch": 6.53371421769922,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 27446
    },
    {
      "epoch": 6.5339522704279,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 27447
    },
    {
      "epoch": 6.534190323156579,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 27448
    },
    {
      "epoch": 6.534428375885259,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 27449
    },
    {
      "epoch": 6.534666428613938,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 27450
    },
    {
      "epoch": 6.534904481342617,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 27451
    },
    {
      "epoch": 6.535142534071297,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 27452
    },
    {
      "epoch": 6.535380586799977,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 27453
    },
    {
      "epoch": 6.535618639528655,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 27454
    },
    {
      "epoch": 6.535856692257335,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 27455
    },
    {
      "epoch": 6.5360947449860145,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 27456
    },
    {
      "epoch": 6.536332797714694,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 27457
    },
    {
      "epoch": 6.536570850443373,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 27458
    },
    {
      "epoch": 6.536808903172052,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 27459
    },
    {
      "epoch": 6.537046955900732,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 27460
    },
    {
      "epoch": 6.537285008629412,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3908,
      "step": 27461
    },
    {
      "epoch": 6.53752306135809,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 27462
    },
    {
      "epoch": 6.53776111408677,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 27463
    },
    {
      "epoch": 6.53799916681545,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 27464
    },
    {
      "epoch": 6.538237219544129,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 27465
    },
    {
      "epoch": 6.538475272272809,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 27466
    },
    {
      "epoch": 6.5387133250014875,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 27467
    },
    {
      "epoch": 6.538951377730167,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 27468
    },
    {
      "epoch": 6.539189430458847,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 27469
    },
    {
      "epoch": 6.539427483187526,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 27470
    },
    {
      "epoch": 6.539665535916205,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3803,
      "step": 27471
    },
    {
      "epoch": 6.539903588644885,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 27472
    },
    {
      "epoch": 6.540141641373564,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.5355,
      "step": 27473
    },
    {
      "epoch": 6.540379694102244,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 27474
    },
    {
      "epoch": 6.5406177468309235,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 27475
    },
    {
      "epoch": 6.540855799559602,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 27476
    },
    {
      "epoch": 6.541093852288282,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 27477
    },
    {
      "epoch": 6.541331905016961,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1668,
      "step": 27478
    },
    {
      "epoch": 6.541569957745641,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.734,
      "step": 27479
    },
    {
      "epoch": 6.54180801047432,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 27480
    },
    {
      "epoch": 6.542046063202999,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2811,
      "step": 27481
    },
    {
      "epoch": 6.542284115931679,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 27482
    },
    {
      "epoch": 6.5425221686603585,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 27483
    },
    {
      "epoch": 6.542760221389038,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.3206,
      "step": 27484
    },
    {
      "epoch": 6.542998274117717,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2917,
      "step": 27485
    },
    {
      "epoch": 6.5432363268463964,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 27486
    },
    {
      "epoch": 6.543474379575076,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 27487
    },
    {
      "epoch": 6.543712432303756,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 27488
    },
    {
      "epoch": 6.543950485032434,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 27489
    },
    {
      "epoch": 6.544188537761114,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 27490
    },
    {
      "epoch": 6.544426590489794,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 27491
    },
    {
      "epoch": 6.544664643218473,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 27492
    },
    {
      "epoch": 6.544902695947153,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 27493
    },
    {
      "epoch": 6.5451407486758315,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 27494
    },
    {
      "epoch": 6.545378801404511,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 27495
    },
    {
      "epoch": 6.545616854133191,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 27496
    },
    {
      "epoch": 6.545854906861869,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 27497
    },
    {
      "epoch": 6.546092959590549,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 27498
    },
    {
      "epoch": 6.546331012319229,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 27499
    },
    {
      "epoch": 6.546569065047908,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 27500
    },
    {
      "epoch": 6.546569065047908,
      "eval_runtime": 0.5252,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 27500
    },
    {
      "epoch": 6.546807117776588,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 27501
    },
    {
      "epoch": 6.547045170505267,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1526,
      "step": 27502
    },
    {
      "epoch": 6.547283223233946,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 27503
    },
    {
      "epoch": 6.547521275962626,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 27504
    },
    {
      "epoch": 6.547759328691305,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 27505
    },
    {
      "epoch": 6.547997381419984,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 27506
    },
    {
      "epoch": 6.548235434148664,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 27507
    },
    {
      "epoch": 6.548473486877343,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 27508
    },
    {
      "epoch": 6.548711539606023,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.6412,
      "step": 27509
    },
    {
      "epoch": 6.5489495923347025,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 27510
    },
    {
      "epoch": 6.549187645063381,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 27511
    },
    {
      "epoch": 6.549425697792061,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4845,
      "step": 27512
    },
    {
      "epoch": 6.5496637505207405,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 27513
    },
    {
      "epoch": 6.54990180324942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 27514
    },
    {
      "epoch": 6.550139855978099,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 27515
    },
    {
      "epoch": 6.550377908706778,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 27516
    },
    {
      "epoch": 6.550615961435458,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 27517
    },
    {
      "epoch": 6.550854014164138,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 27518
    },
    {
      "epoch": 6.551092066892817,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1747,
      "step": 27519
    },
    {
      "epoch": 6.551330119621496,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 27520
    },
    {
      "epoch": 6.5515681723501755,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 27521
    },
    {
      "epoch": 6.551806225078855,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 27522
    },
    {
      "epoch": 6.552044277807535,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 27523
    },
    {
      "epoch": 6.5522823305362135,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2847,
      "step": 27524
    },
    {
      "epoch": 6.552520383264893,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 27525
    },
    {
      "epoch": 6.552758435993573,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 27526
    },
    {
      "epoch": 6.552996488722252,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 27527
    },
    {
      "epoch": 6.553234541450932,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8822,
      "step": 27528
    },
    {
      "epoch": 6.553472594179611,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 27529
    },
    {
      "epoch": 6.55371064690829,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 27530
    },
    {
      "epoch": 6.55394869963697,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 27531
    },
    {
      "epoch": 6.5541867523656485,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 27532
    },
    {
      "epoch": 6.554424805094328,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 27533
    },
    {
      "epoch": 6.554662857823008,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 27534
    },
    {
      "epoch": 6.554900910551687,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 27535
    },
    {
      "epoch": 6.555138963280367,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 27536
    },
    {
      "epoch": 6.555377016009046,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 27537
    },
    {
      "epoch": 6.555615068737725,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 27538
    },
    {
      "epoch": 6.555853121466405,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 27539
    },
    {
      "epoch": 6.5560911741950845,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 27540
    },
    {
      "epoch": 6.556329226923763,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 27541
    },
    {
      "epoch": 6.556567279652443,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9604,
      "step": 27542
    },
    {
      "epoch": 6.556805332381122,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.434,
      "step": 27543
    },
    {
      "epoch": 6.557043385109802,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 27544
    },
    {
      "epoch": 6.557281437838482,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 27545
    },
    {
      "epoch": 6.55751949056716,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3463,
      "step": 27546
    },
    {
      "epoch": 6.55775754329584,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 27547
    },
    {
      "epoch": 6.5579955960245195,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 27548
    },
    {
      "epoch": 6.558233648753199,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 27549
    },
    {
      "epoch": 6.558471701481878,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.739,
      "step": 27550
    },
    {
      "epoch": 6.5587097542105575,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0925,
      "step": 27551
    },
    {
      "epoch": 6.558947806939237,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 27552
    },
    {
      "epoch": 6.559185859667917,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 27553
    },
    {
      "epoch": 6.559423912396596,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 27554
    },
    {
      "epoch": 6.559661965125275,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 27555
    },
    {
      "epoch": 6.559900017853955,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 27556
    },
    {
      "epoch": 6.560138070582634,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 27557
    },
    {
      "epoch": 6.560376123311314,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 27558
    },
    {
      "epoch": 6.5606141760399925,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 27559
    },
    {
      "epoch": 6.560852228768672,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 27560
    },
    {
      "epoch": 6.561090281497352,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 27561
    },
    {
      "epoch": 6.561328334226031,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 27562
    },
    {
      "epoch": 6.561566386954711,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 27563
    },
    {
      "epoch": 6.56180443968339,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 27564
    },
    {
      "epoch": 6.562042492412069,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 27565
    },
    {
      "epoch": 6.562280545140749,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 27566
    },
    {
      "epoch": 6.562518597869428,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 27567
    },
    {
      "epoch": 6.562756650598107,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8468,
      "step": 27568
    },
    {
      "epoch": 6.562994703326787,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 27569
    },
    {
      "epoch": 6.563232756055466,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 27570
    },
    {
      "epoch": 6.563470808784146,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 27571
    },
    {
      "epoch": 6.563708861512825,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 27572
    },
    {
      "epoch": 6.563946914241504,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 27573
    },
    {
      "epoch": 6.564184966970184,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 27574
    },
    {
      "epoch": 6.5644230196988635,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 27575
    },
    {
      "epoch": 6.564661072427542,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 27576
    },
    {
      "epoch": 6.564899125156222,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 27577
    },
    {
      "epoch": 6.5651371778849015,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 27578
    },
    {
      "epoch": 6.565375230613581,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 27579
    },
    {
      "epoch": 6.565613283342261,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 27580
    },
    {
      "epoch": 6.565851336070939,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 27581
    },
    {
      "epoch": 6.566089388799619,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 27582
    },
    {
      "epoch": 6.566327441528299,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 27583
    },
    {
      "epoch": 6.566565494256978,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 27584
    },
    {
      "epoch": 6.566803546985657,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.311,
      "step": 27585
    },
    {
      "epoch": 6.5670415997143365,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 27586
    },
    {
      "epoch": 6.567279652443016,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8421,
      "step": 27587
    },
    {
      "epoch": 6.567517705171696,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 27588
    },
    {
      "epoch": 6.567755757900375,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.945,
      "step": 27589
    },
    {
      "epoch": 6.567993810629054,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 27590
    },
    {
      "epoch": 6.568231863357734,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 27591
    },
    {
      "epoch": 6.568469916086413,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 27592
    },
    {
      "epoch": 6.568707968815093,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2922,
      "step": 27593
    },
    {
      "epoch": 6.568946021543772,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 27594
    },
    {
      "epoch": 6.569184074272451,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9738,
      "step": 27595
    },
    {
      "epoch": 6.569422127001131,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 27596
    },
    {
      "epoch": 6.56966017972981,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 27597
    },
    {
      "epoch": 6.56989823245849,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 27598
    },
    {
      "epoch": 6.570136285187169,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 27599
    },
    {
      "epoch": 6.570374337915848,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0613,
      "step": 27600
    },
    {
      "epoch": 6.570374337915848,
      "eval_runtime": 0.5345,
      "eval_samples_per_second": 1.871,
      "eval_steps_per_second": 1.871,
      "step": 27600
    },
    {
      "epoch": 6.570612390644528,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3459,
      "step": 27601
    },
    {
      "epoch": 6.570850443373207,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2606,
      "step": 27602
    },
    {
      "epoch": 6.571088496101886,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 27603
    },
    {
      "epoch": 6.571326548830566,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 27604
    },
    {
      "epoch": 6.5715646015592455,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 27605
    },
    {
      "epoch": 6.571802654287925,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 27606
    },
    {
      "epoch": 6.572040707016604,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3701,
      "step": 27607
    },
    {
      "epoch": 6.572278759745283,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 27608
    },
    {
      "epoch": 6.572516812473963,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.6751,
      "step": 27609
    },
    {
      "epoch": 6.572754865202643,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 27610
    },
    {
      "epoch": 6.572992917931321,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 27611
    },
    {
      "epoch": 6.573230970660001,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 27612
    },
    {
      "epoch": 6.5734690233886806,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 27613
    },
    {
      "epoch": 6.57370707611736,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 27614
    },
    {
      "epoch": 6.57394512884604,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 27615
    },
    {
      "epoch": 6.5741831815747185,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 27616
    },
    {
      "epoch": 6.574421234303398,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 27617
    },
    {
      "epoch": 6.574659287032078,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8607,
      "step": 27618
    },
    {
      "epoch": 6.574897339760757,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9267,
      "step": 27619
    },
    {
      "epoch": 6.575135392489436,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3657,
      "step": 27620
    },
    {
      "epoch": 6.575373445218116,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 27621
    },
    {
      "epoch": 6.575611497946795,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 27622
    },
    {
      "epoch": 6.575849550675475,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 27623
    },
    {
      "epoch": 6.576087603404154,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 27624
    },
    {
      "epoch": 6.576325656132833,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 27625
    },
    {
      "epoch": 6.576563708861513,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 27626
    },
    {
      "epoch": 6.576801761590192,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 27627
    },
    {
      "epoch": 6.577039814318872,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 27628
    },
    {
      "epoch": 6.577277867047551,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 27629
    },
    {
      "epoch": 6.57751591977623,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 27630
    },
    {
      "epoch": 6.57775397250491,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 27631
    },
    {
      "epoch": 6.5779920252335895,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 27632
    },
    {
      "epoch": 6.578230077962269,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 27633
    },
    {
      "epoch": 6.578468130690948,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 27634
    },
    {
      "epoch": 6.578706183419627,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 27635
    },
    {
      "epoch": 6.578944236148307,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.481,
      "step": 27636
    },
    {
      "epoch": 6.579182288876987,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 27637
    },
    {
      "epoch": 6.579420341605665,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 27638
    },
    {
      "epoch": 6.579658394334345,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 27639
    },
    {
      "epoch": 6.579896447063025,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3715,
      "step": 27640
    },
    {
      "epoch": 6.580134499791704,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 27641
    },
    {
      "epoch": 6.580372552520383,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 27642
    },
    {
      "epoch": 6.5806106052490625,
      "grad_norm": 2.703125,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 27643
    },
    {
      "epoch": 6.580848657977742,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 27644
    },
    {
      "epoch": 6.581086710706422,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 27645
    },
    {
      "epoch": 6.5813247634351,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 27646
    },
    {
      "epoch": 6.58156281616378,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 27647
    },
    {
      "epoch": 6.58180086889246,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 27648
    },
    {
      "epoch": 6.582038921621139,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 27649
    },
    {
      "epoch": 6.582276974349819,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 27650
    },
    {
      "epoch": 6.582515027078498,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 27651
    },
    {
      "epoch": 6.582753079807177,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 27652
    },
    {
      "epoch": 6.582991132535857,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 27653
    },
    {
      "epoch": 6.583229185264536,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 27654
    },
    {
      "epoch": 6.583467237993215,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 27655
    },
    {
      "epoch": 6.583705290721895,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.4028,
      "step": 27656
    },
    {
      "epoch": 6.583943343450574,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 27657
    },
    {
      "epoch": 6.584181396179254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 27658
    },
    {
      "epoch": 6.5844194489079335,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 27659
    },
    {
      "epoch": 6.584657501636612,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 27660
    },
    {
      "epoch": 6.584895554365292,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 27661
    },
    {
      "epoch": 6.585133607093971,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.6979,
      "step": 27662
    },
    {
      "epoch": 6.585371659822651,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 27663
    },
    {
      "epoch": 6.58560971255133,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 27664
    },
    {
      "epoch": 6.585847765280009,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 27665
    },
    {
      "epoch": 6.586085818008689,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2438,
      "step": 27666
    },
    {
      "epoch": 6.586323870737369,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 27667
    },
    {
      "epoch": 6.586561923466048,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7727,
      "step": 27668
    },
    {
      "epoch": 6.586799976194727,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 27669
    },
    {
      "epoch": 6.5870380289234065,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 27670
    },
    {
      "epoch": 6.587276081652086,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 27671
    },
    {
      "epoch": 6.587514134380766,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 27672
    },
    {
      "epoch": 6.587752187109444,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 27673
    },
    {
      "epoch": 6.587990239838124,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 27674
    },
    {
      "epoch": 6.588228292566804,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 27675
    },
    {
      "epoch": 6.588466345295483,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 27676
    },
    {
      "epoch": 6.588704398024162,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 27677
    },
    {
      "epoch": 6.588942450752842,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 27678
    },
    {
      "epoch": 6.589180503481521,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 27679
    },
    {
      "epoch": 6.589418556210201,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 27680
    },
    {
      "epoch": 6.5896566089388795,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0159,
      "step": 27681
    },
    {
      "epoch": 6.589894661667559,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 27682
    },
    {
      "epoch": 6.590132714396239,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 27683
    },
    {
      "epoch": 6.590370767124918,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 27684
    },
    {
      "epoch": 6.590608819853598,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2123,
      "step": 27685
    },
    {
      "epoch": 6.590846872582277,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 27686
    },
    {
      "epoch": 6.591084925310956,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.7769,
      "step": 27687
    },
    {
      "epoch": 6.591322978039636,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 27688
    },
    {
      "epoch": 6.5915610307683155,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 27689
    },
    {
      "epoch": 6.591799083496994,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 27690
    },
    {
      "epoch": 6.592037136225674,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 27691
    },
    {
      "epoch": 6.592275188954353,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9158,
      "step": 27692
    },
    {
      "epoch": 6.592513241683033,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 27693
    },
    {
      "epoch": 6.592751294411713,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 27694
    },
    {
      "epoch": 6.592989347140391,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 27695
    },
    {
      "epoch": 6.593227399869071,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1509,
      "step": 27696
    },
    {
      "epoch": 6.5934654525977505,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 27697
    },
    {
      "epoch": 6.59370350532643,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 27698
    },
    {
      "epoch": 6.593941558055109,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 27699
    },
    {
      "epoch": 6.5941796107837884,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 27700
    },
    {
      "epoch": 6.5941796107837884,
      "eval_runtime": 0.5305,
      "eval_samples_per_second": 1.885,
      "eval_steps_per_second": 1.885,
      "step": 27700
    },
    {
      "epoch": 6.594417663512468,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 27701
    },
    {
      "epoch": 6.594655716241148,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 27702
    },
    {
      "epoch": 6.594893768969827,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4655,
      "step": 27703
    },
    {
      "epoch": 6.595131821698506,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 27704
    },
    {
      "epoch": 6.595369874427186,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 27705
    },
    {
      "epoch": 6.595607927155865,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.4246,
      "step": 27706
    },
    {
      "epoch": 6.595845979884545,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 27707
    },
    {
      "epoch": 6.5960840326132235,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 27708
    },
    {
      "epoch": 6.596322085341903,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7942,
      "step": 27709
    },
    {
      "epoch": 6.596560138070583,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 27710
    },
    {
      "epoch": 6.596798190799262,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 27711
    },
    {
      "epoch": 6.597036243527942,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 27712
    },
    {
      "epoch": 6.597274296256621,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 27713
    },
    {
      "epoch": 6.5975123489853,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 27714
    },
    {
      "epoch": 6.59775040171398,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4176,
      "step": 27715
    },
    {
      "epoch": 6.597988454442659,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 27716
    },
    {
      "epoch": 6.598226507171338,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.135,
      "step": 27717
    },
    {
      "epoch": 6.598464559900018,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 27718
    },
    {
      "epoch": 6.598702612628697,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 27719
    },
    {
      "epoch": 6.598940665357377,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 27720
    },
    {
      "epoch": 6.599178718086056,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.116,
      "step": 27721
    },
    {
      "epoch": 6.599416770814735,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 27722
    },
    {
      "epoch": 6.599654823543415,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2612,
      "step": 27723
    },
    {
      "epoch": 6.5998928762720945,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8237,
      "step": 27724
    },
    {
      "epoch": 6.600130929000773,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 27725
    },
    {
      "epoch": 6.600368981729453,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2631,
      "step": 27726
    },
    {
      "epoch": 6.6006070344581325,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 27727
    },
    {
      "epoch": 6.600845087186812,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 27728
    },
    {
      "epoch": 6.601083139915492,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 27729
    },
    {
      "epoch": 6.60132119264417,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 27730
    },
    {
      "epoch": 6.60155924537285,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 27731
    },
    {
      "epoch": 6.60179729810153,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0181,
      "step": 27732
    },
    {
      "epoch": 6.602035350830209,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 27733
    },
    {
      "epoch": 6.602273403558888,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3767,
      "step": 27734
    },
    {
      "epoch": 6.6025114562875675,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 27735
    },
    {
      "epoch": 6.602749509016247,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2619,
      "step": 27736
    },
    {
      "epoch": 6.602987561744927,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 27737
    },
    {
      "epoch": 6.603225614473606,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 27738
    },
    {
      "epoch": 6.603463667202285,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 27739
    },
    {
      "epoch": 6.603701719930965,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 27740
    },
    {
      "epoch": 6.603939772659644,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 27741
    },
    {
      "epoch": 6.604177825388324,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 27742
    },
    {
      "epoch": 6.604415878117003,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 27743
    },
    {
      "epoch": 6.604653930845682,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 27744
    },
    {
      "epoch": 6.604891983574362,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 27745
    },
    {
      "epoch": 6.605130036303041,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 27746
    },
    {
      "epoch": 6.605368089031721,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 27747
    },
    {
      "epoch": 6.6056061417604,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 27748
    },
    {
      "epoch": 6.605844194489079,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 27749
    },
    {
      "epoch": 6.606082247217759,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 27750
    },
    {
      "epoch": 6.606320299946438,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3136,
      "step": 27751
    },
    {
      "epoch": 6.606558352675117,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 27752
    },
    {
      "epoch": 6.606796405403797,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 27753
    },
    {
      "epoch": 6.6070344581324765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 27754
    },
    {
      "epoch": 6.607272510861156,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 27755
    },
    {
      "epoch": 6.607510563589835,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 27756
    },
    {
      "epoch": 6.607748616318514,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 27757
    },
    {
      "epoch": 6.607986669047194,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 27758
    },
    {
      "epoch": 6.608224721775874,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 27759
    },
    {
      "epoch": 6.608462774504552,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 27760
    },
    {
      "epoch": 6.608700827233232,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 27761
    },
    {
      "epoch": 6.6089388799619115,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 27762
    },
    {
      "epoch": 6.609176932690591,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 27763
    },
    {
      "epoch": 6.609414985419271,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 27764
    },
    {
      "epoch": 6.6096530381479495,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8935,
      "step": 27765
    },
    {
      "epoch": 6.609891090876629,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 27766
    },
    {
      "epoch": 6.610129143605309,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 27767
    },
    {
      "epoch": 6.610367196333988,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 27768
    },
    {
      "epoch": 6.610605249062667,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9028,
      "step": 27769
    },
    {
      "epoch": 6.610843301791347,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 27770
    },
    {
      "epoch": 6.611081354520026,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 27771
    },
    {
      "epoch": 6.611319407248706,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 27772
    },
    {
      "epoch": 6.611557459977385,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 27773
    },
    {
      "epoch": 6.611795512706064,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 27774
    },
    {
      "epoch": 6.612033565434744,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 27775
    },
    {
      "epoch": 6.612271618163423,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 27776
    },
    {
      "epoch": 6.612509670892103,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 27777
    },
    {
      "epoch": 6.612747723620782,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 27778
    },
    {
      "epoch": 6.612985776349461,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 27779
    },
    {
      "epoch": 6.613223829078141,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 27780
    },
    {
      "epoch": 6.6134618818068205,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 27781
    },
    {
      "epoch": 6.6136999345355,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 27782
    },
    {
      "epoch": 6.613937987264179,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 27783
    },
    {
      "epoch": 6.614176039992858,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 27784
    },
    {
      "epoch": 6.614414092721538,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 27785
    },
    {
      "epoch": 6.614652145450217,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.6313,
      "step": 27786
    },
    {
      "epoch": 6.614890198178896,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8819,
      "step": 27787
    },
    {
      "epoch": 6.615128250907576,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0869,
      "step": 27788
    },
    {
      "epoch": 6.6153663036362556,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 27789
    },
    {
      "epoch": 6.615604356364935,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 27790
    },
    {
      "epoch": 6.615842409093614,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 27791
    },
    {
      "epoch": 6.6160804618222935,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 27792
    },
    {
      "epoch": 6.616318514550973,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 27793
    },
    {
      "epoch": 6.616556567279653,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 27794
    },
    {
      "epoch": 6.616794620008331,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 27795
    },
    {
      "epoch": 6.617032672737011,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.765,
      "step": 27796
    },
    {
      "epoch": 6.617270725465691,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3825,
      "step": 27797
    },
    {
      "epoch": 6.61750877819437,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 27798
    },
    {
      "epoch": 6.61774683092305,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2631,
      "step": 27799
    },
    {
      "epoch": 6.6179848836517285,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 27800
    },
    {
      "epoch": 6.6179848836517285,
      "eval_runtime": 0.5247,
      "eval_samples_per_second": 1.906,
      "eval_steps_per_second": 1.906,
      "step": 27800
    },
    {
      "epoch": 6.618222936380408,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 27801
    },
    {
      "epoch": 6.618460989109088,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 27802
    },
    {
      "epoch": 6.618699041837767,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 27803
    },
    {
      "epoch": 6.618937094566446,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 27804
    },
    {
      "epoch": 6.619175147295126,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 27805
    },
    {
      "epoch": 6.619413200023805,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.417,
      "step": 27806
    },
    {
      "epoch": 6.619651252752485,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 27807
    },
    {
      "epoch": 6.6198893054811645,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4228,
      "step": 27808
    },
    {
      "epoch": 6.620127358209843,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 27809
    },
    {
      "epoch": 6.620365410938523,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 27810
    },
    {
      "epoch": 6.620603463667202,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1771,
      "step": 27811
    },
    {
      "epoch": 6.620841516395882,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 27812
    },
    {
      "epoch": 6.621079569124561,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 27813
    },
    {
      "epoch": 6.62131762185324,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 27814
    },
    {
      "epoch": 6.62155567458192,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 27815
    },
    {
      "epoch": 6.6217937273106,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3213,
      "step": 27816
    },
    {
      "epoch": 6.622031780039279,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 27817
    },
    {
      "epoch": 6.622269832767958,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1037,
      "step": 27818
    },
    {
      "epoch": 6.6225078854966375,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 27819
    },
    {
      "epoch": 6.622745938225317,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 27820
    },
    {
      "epoch": 6.622983990953996,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 27821
    },
    {
      "epoch": 6.623222043682675,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 27822
    },
    {
      "epoch": 6.623460096411355,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 27823
    },
    {
      "epoch": 6.623698149140035,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 27824
    },
    {
      "epoch": 6.623936201868714,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 27825
    },
    {
      "epoch": 6.624174254597393,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 27826
    },
    {
      "epoch": 6.624412307326073,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 27827
    },
    {
      "epoch": 6.624650360054752,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 27828
    },
    {
      "epoch": 6.624888412783432,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 27829
    },
    {
      "epoch": 6.6251264655121105,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 27830
    },
    {
      "epoch": 6.62536451824079,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 27831
    },
    {
      "epoch": 6.62560257096947,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 27832
    },
    {
      "epoch": 6.625840623698149,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 27833
    },
    {
      "epoch": 6.626078676426829,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 27834
    },
    {
      "epoch": 6.626316729155508,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 27835
    },
    {
      "epoch": 6.626554781884187,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 27836
    },
    {
      "epoch": 6.626792834612867,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 27837
    },
    {
      "epoch": 6.627030887341546,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 27838
    },
    {
      "epoch": 6.627268940070225,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 27839
    },
    {
      "epoch": 6.627506992798905,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.3417,
      "step": 27840
    },
    {
      "epoch": 6.627745045527584,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 27841
    },
    {
      "epoch": 6.627983098256264,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2565,
      "step": 27842
    },
    {
      "epoch": 6.628221150984944,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 27843
    },
    {
      "epoch": 6.628459203713622,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 27844
    },
    {
      "epoch": 6.628697256442302,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4465,
      "step": 27845
    },
    {
      "epoch": 6.6289353091709815,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.5101,
      "step": 27846
    },
    {
      "epoch": 6.629173361899661,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 27847
    },
    {
      "epoch": 6.62941141462834,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 27848
    },
    {
      "epoch": 6.629649467357019,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 27849
    },
    {
      "epoch": 6.629887520085699,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7199,
      "step": 27850
    },
    {
      "epoch": 6.630125572814379,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 27851
    },
    {
      "epoch": 6.630363625543058,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 27852
    },
    {
      "epoch": 6.630601678271737,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 27853
    },
    {
      "epoch": 6.630839731000417,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 27854
    },
    {
      "epoch": 6.631077783729096,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 27855
    },
    {
      "epoch": 6.631315836457776,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 27856
    },
    {
      "epoch": 6.6315538891864545,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 27857
    },
    {
      "epoch": 6.631791941915134,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3735,
      "step": 27858
    },
    {
      "epoch": 6.632029994643814,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 27859
    },
    {
      "epoch": 6.632268047372493,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 27860
    },
    {
      "epoch": 6.632506100101172,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 27861
    },
    {
      "epoch": 6.632744152829852,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 27862
    },
    {
      "epoch": 6.632982205558531,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1485,
      "step": 27863
    },
    {
      "epoch": 6.633220258287211,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 27864
    },
    {
      "epoch": 6.63345831101589,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 27865
    },
    {
      "epoch": 6.633696363744569,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 27866
    },
    {
      "epoch": 6.633934416473249,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 27867
    },
    {
      "epoch": 6.634172469201928,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 27868
    },
    {
      "epoch": 6.634410521930608,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 27869
    },
    {
      "epoch": 6.634648574659287,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.757,
      "step": 27870
    },
    {
      "epoch": 6.634886627387966,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 27871
    },
    {
      "epoch": 6.635124680116646,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 27872
    },
    {
      "epoch": 6.6353627328453255,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 27873
    },
    {
      "epoch": 6.635600785574004,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 27874
    },
    {
      "epoch": 6.635838838302684,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 27875
    },
    {
      "epoch": 6.6360768910313634,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 27876
    },
    {
      "epoch": 6.636314943760043,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4124,
      "step": 27877
    },
    {
      "epoch": 6.636552996488723,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 27878
    },
    {
      "epoch": 6.636791049217401,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 27879
    },
    {
      "epoch": 6.637029101946081,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 27880
    },
    {
      "epoch": 6.637267154674761,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 27881
    },
    {
      "epoch": 6.63750520740344,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 27882
    },
    {
      "epoch": 6.637743260132119,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 27883
    },
    {
      "epoch": 6.6379813128607985,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4071,
      "step": 27884
    },
    {
      "epoch": 6.638219365589478,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 27885
    },
    {
      "epoch": 6.638457418318158,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 27886
    },
    {
      "epoch": 6.638695471046837,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 27887
    },
    {
      "epoch": 6.638933523775516,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3497,
      "step": 27888
    },
    {
      "epoch": 6.639171576504196,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 27889
    },
    {
      "epoch": 6.639409629232875,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 27890
    },
    {
      "epoch": 6.639647681961555,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2675,
      "step": 27891
    },
    {
      "epoch": 6.639885734690234,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 27892
    },
    {
      "epoch": 6.640123787418913,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1882,
      "step": 27893
    },
    {
      "epoch": 6.640361840147593,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 27894
    },
    {
      "epoch": 6.640599892876272,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 27895
    },
    {
      "epoch": 6.640837945604951,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 27896
    },
    {
      "epoch": 6.641075998333631,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 27897
    },
    {
      "epoch": 6.64131405106231,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 27898
    },
    {
      "epoch": 6.64155210379099,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 27899
    },
    {
      "epoch": 6.641790156519669,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 27900
    },
    {
      "epoch": 6.641790156519669,
      "eval_runtime": 0.5204,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 27900
    },
    {
      "epoch": 6.642028209248348,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 27901
    },
    {
      "epoch": 6.642266261977028,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 27902
    },
    {
      "epoch": 6.6425043147057075,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 27903
    },
    {
      "epoch": 6.642742367434387,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 27904
    },
    {
      "epoch": 6.642980420163066,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 27905
    },
    {
      "epoch": 6.643218472891745,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 27906
    },
    {
      "epoch": 6.643456525620425,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9249,
      "step": 27907
    },
    {
      "epoch": 6.643694578349105,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 27908
    },
    {
      "epoch": 6.643932631077783,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4521,
      "step": 27909
    },
    {
      "epoch": 6.644170683806463,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 27910
    },
    {
      "epoch": 6.6444087365351425,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1649,
      "step": 27911
    },
    {
      "epoch": 6.644646789263822,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 27912
    },
    {
      "epoch": 6.644884841992502,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.369,
      "step": 27913
    },
    {
      "epoch": 6.6451228947211805,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 27914
    },
    {
      "epoch": 6.64536094744986,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2144,
      "step": 27915
    },
    {
      "epoch": 6.64559900017854,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 27916
    },
    {
      "epoch": 6.645837052907219,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 27917
    },
    {
      "epoch": 6.646075105635898,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 27918
    },
    {
      "epoch": 6.646313158364578,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 27919
    },
    {
      "epoch": 6.646551211093257,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 27920
    },
    {
      "epoch": 6.646789263821937,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 27921
    },
    {
      "epoch": 6.647027316550616,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 27922
    },
    {
      "epoch": 6.647265369279295,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 27923
    },
    {
      "epoch": 6.647503422007975,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 27924
    },
    {
      "epoch": 6.647741474736654,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 27925
    },
    {
      "epoch": 6.647979527465334,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 27926
    },
    {
      "epoch": 6.648217580194013,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 27927
    },
    {
      "epoch": 6.648455632922692,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 27928
    },
    {
      "epoch": 6.648693685651372,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 27929
    },
    {
      "epoch": 6.6489317383800515,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 27930
    },
    {
      "epoch": 6.649169791108731,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 27931
    },
    {
      "epoch": 6.64940784383741,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 27932
    },
    {
      "epoch": 6.649645896566089,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 27933
    },
    {
      "epoch": 6.649883949294769,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 27934
    },
    {
      "epoch": 6.650122002023448,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 27935
    },
    {
      "epoch": 6.650360054752127,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 27936
    },
    {
      "epoch": 6.650598107480807,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 27937
    },
    {
      "epoch": 6.6508361602094865,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3268,
      "step": 27938
    },
    {
      "epoch": 6.651074212938166,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 27939
    },
    {
      "epoch": 6.651312265666845,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 27940
    },
    {
      "epoch": 6.6515503183955245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 27941
    },
    {
      "epoch": 6.651788371124204,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 27942
    },
    {
      "epoch": 6.652026423852884,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 27943
    },
    {
      "epoch": 6.652264476581562,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 27944
    },
    {
      "epoch": 6.652502529310242,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 27945
    },
    {
      "epoch": 6.652740582038922,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 27946
    },
    {
      "epoch": 6.652978634767601,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 27947
    },
    {
      "epoch": 6.653216687496281,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 27948
    },
    {
      "epoch": 6.6534547402249595,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 27949
    },
    {
      "epoch": 6.653692792953639,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1993,
      "step": 27950
    },
    {
      "epoch": 6.653930845682319,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 27951
    },
    {
      "epoch": 6.654168898410998,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 27952
    },
    {
      "epoch": 6.654406951139677,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 27953
    },
    {
      "epoch": 6.654645003868357,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 27954
    },
    {
      "epoch": 6.654883056597036,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6964,
      "step": 27955
    },
    {
      "epoch": 6.655121109325716,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 27956
    },
    {
      "epoch": 6.6553591620543955,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 27957
    },
    {
      "epoch": 6.655597214783074,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 27958
    },
    {
      "epoch": 6.655835267511754,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 27959
    },
    {
      "epoch": 6.656073320240433,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 27960
    },
    {
      "epoch": 6.656311372969113,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 27961
    },
    {
      "epoch": 6.656549425697792,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 27962
    },
    {
      "epoch": 6.656787478426471,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 27963
    },
    {
      "epoch": 6.657025531155151,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 27964
    },
    {
      "epoch": 6.6572635838838305,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 27965
    },
    {
      "epoch": 6.65750163661251,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 27966
    },
    {
      "epoch": 6.657739689341189,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 27967
    },
    {
      "epoch": 6.6579777420698685,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 27968
    },
    {
      "epoch": 6.658215794798548,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 27969
    },
    {
      "epoch": 6.658453847527227,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 27970
    },
    {
      "epoch": 6.658691900255906,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.291,
      "step": 27971
    },
    {
      "epoch": 6.658929952984586,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 27972
    },
    {
      "epoch": 6.659168005713266,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 27973
    },
    {
      "epoch": 6.659406058441945,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 27974
    },
    {
      "epoch": 6.659644111170624,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 27975
    },
    {
      "epoch": 6.6598821638993035,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1425,
      "step": 27976
    },
    {
      "epoch": 6.660120216627983,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 27977
    },
    {
      "epoch": 6.660358269356663,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 27978
    },
    {
      "epoch": 6.6605963220853415,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 27979
    },
    {
      "epoch": 6.660834374814021,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9954,
      "step": 27980
    },
    {
      "epoch": 6.661072427542701,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4385,
      "step": 27981
    },
    {
      "epoch": 6.66131048027138,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 27982
    },
    {
      "epoch": 6.66154853300006,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 27983
    },
    {
      "epoch": 6.661786585728739,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 27984
    },
    {
      "epoch": 6.662024638457418,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3413,
      "step": 27985
    },
    {
      "epoch": 6.662262691186098,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 27986
    },
    {
      "epoch": 6.662500743914777,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 27987
    },
    {
      "epoch": 6.662738796643456,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 27988
    },
    {
      "epoch": 6.662976849372136,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 27989
    },
    {
      "epoch": 6.663214902100815,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 27990
    },
    {
      "epoch": 6.663452954829495,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4239,
      "step": 27991
    },
    {
      "epoch": 6.663691007558175,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 27992
    },
    {
      "epoch": 6.663929060286853,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 27993
    },
    {
      "epoch": 6.664167113015533,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 27994
    },
    {
      "epoch": 6.6644051657442125,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 27995
    },
    {
      "epoch": 6.664643218472892,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 27996
    },
    {
      "epoch": 6.664881271201571,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 27997
    },
    {
      "epoch": 6.66511932393025,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 27998
    },
    {
      "epoch": 6.66535737665893,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 27999
    },
    {
      "epoch": 6.66559542938761,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 28000
    },
    {
      "epoch": 6.66559542938761,
      "eval_runtime": 0.5335,
      "eval_samples_per_second": 1.875,
      "eval_steps_per_second": 1.875,
      "step": 28000
    },
    {
      "epoch": 6.665833482116289,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 28001
    },
    {
      "epoch": 6.666071534844968,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 28002
    },
    {
      "epoch": 6.6663095875736476,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3511,
      "step": 28003
    },
    {
      "epoch": 6.666547640302327,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 28004
    },
    {
      "epoch": 6.666785693031006,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 28005
    },
    {
      "epoch": 6.6670237457596855,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 28006
    },
    {
      "epoch": 6.667261798488365,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3517,
      "step": 28007
    },
    {
      "epoch": 6.667499851217045,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 28008
    },
    {
      "epoch": 6.667737903945724,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.3356,
      "step": 28009
    },
    {
      "epoch": 6.667975956674403,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 28010
    },
    {
      "epoch": 6.668214009403083,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 28011
    },
    {
      "epoch": 6.668452062131762,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 28012
    },
    {
      "epoch": 6.668690114860442,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 28013
    },
    {
      "epoch": 6.6689281675891205,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 28014
    },
    {
      "epoch": 6.6691662203178,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 28015
    },
    {
      "epoch": 6.66940427304648,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 28016
    },
    {
      "epoch": 6.669642325775159,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8438,
      "step": 28017
    },
    {
      "epoch": 6.669880378503839,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 28018
    },
    {
      "epoch": 6.670118431232518,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 28019
    },
    {
      "epoch": 6.670356483961197,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 28020
    },
    {
      "epoch": 6.670594536689877,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 28021
    },
    {
      "epoch": 6.6708325894185565,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2736,
      "step": 28022
    },
    {
      "epoch": 6.671070642147235,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 28023
    },
    {
      "epoch": 6.671308694875915,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 28024
    },
    {
      "epoch": 6.671546747604594,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 28025
    },
    {
      "epoch": 6.671784800333274,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.5399,
      "step": 28026
    },
    {
      "epoch": 6.672022853061954,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3881,
      "step": 28027
    },
    {
      "epoch": 6.672260905790632,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 28028
    },
    {
      "epoch": 6.672498958519312,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 28029
    },
    {
      "epoch": 6.672737011247992,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 28030
    },
    {
      "epoch": 6.672975063976671,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 28031
    },
    {
      "epoch": 6.67321311670535,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 28032
    },
    {
      "epoch": 6.6734511694340295,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7562,
      "step": 28033
    },
    {
      "epoch": 6.673689222162709,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0032,
      "step": 28034
    },
    {
      "epoch": 6.673927274891389,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 28035
    },
    {
      "epoch": 6.674165327620068,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 28036
    },
    {
      "epoch": 6.674403380348747,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9485,
      "step": 28037
    },
    {
      "epoch": 6.674641433077427,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 28038
    },
    {
      "epoch": 6.674879485806106,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 28039
    },
    {
      "epoch": 6.675117538534785,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 28040
    },
    {
      "epoch": 6.675355591263465,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 28041
    },
    {
      "epoch": 6.675593643992144,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8039,
      "step": 28042
    },
    {
      "epoch": 6.675831696720824,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 28043
    },
    {
      "epoch": 6.676069749449503,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 28044
    },
    {
      "epoch": 6.676307802178182,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7285,
      "step": 28045
    },
    {
      "epoch": 6.676545854906862,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 28046
    },
    {
      "epoch": 6.676783907635541,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 28047
    },
    {
      "epoch": 6.677021960364221,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 28048
    },
    {
      "epoch": 6.6772600130929,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9337,
      "step": 28049
    },
    {
      "epoch": 6.677498065821579,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 28050
    },
    {
      "epoch": 6.677736118550259,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 28051
    },
    {
      "epoch": 6.677974171278938,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 28052
    },
    {
      "epoch": 6.678212224007618,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 28053
    },
    {
      "epoch": 6.678450276736297,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7291,
      "step": 28054
    },
    {
      "epoch": 6.678688329464976,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 28055
    },
    {
      "epoch": 6.678926382193656,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 28056
    },
    {
      "epoch": 6.679164434922336,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 28057
    },
    {
      "epoch": 6.679402487651014,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 28058
    },
    {
      "epoch": 6.679640540379694,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 28059
    },
    {
      "epoch": 6.6798785931083735,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 28060
    },
    {
      "epoch": 6.680116645837053,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 28061
    },
    {
      "epoch": 6.680354698565733,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 28062
    },
    {
      "epoch": 6.680592751294411,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 28063
    },
    {
      "epoch": 6.680830804023091,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 28064
    },
    {
      "epoch": 6.681068856751771,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 28065
    },
    {
      "epoch": 6.68130690948045,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3111,
      "step": 28066
    },
    {
      "epoch": 6.681544962209129,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9089,
      "step": 28067
    },
    {
      "epoch": 6.681783014937809,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 28068
    },
    {
      "epoch": 6.682021067666488,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 28069
    },
    {
      "epoch": 6.682259120395168,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 28070
    },
    {
      "epoch": 6.682497173123847,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.4869,
      "step": 28071
    },
    {
      "epoch": 6.682735225852526,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 28072
    },
    {
      "epoch": 6.682973278581206,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 28073
    },
    {
      "epoch": 6.683211331309885,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3186,
      "step": 28074
    },
    {
      "epoch": 6.683449384038565,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 28075
    },
    {
      "epoch": 6.683687436767244,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 28076
    },
    {
      "epoch": 6.683925489495923,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 28077
    },
    {
      "epoch": 6.684163542224603,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 28078
    },
    {
      "epoch": 6.6844015949532825,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 28079
    },
    {
      "epoch": 6.684639647681961,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 28080
    },
    {
      "epoch": 6.684877700410641,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 28081
    },
    {
      "epoch": 6.68511575313932,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 28082
    },
    {
      "epoch": 6.685353805868,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 28083
    },
    {
      "epoch": 6.685591858596679,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 28084
    },
    {
      "epoch": 6.685829911325358,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 28085
    },
    {
      "epoch": 6.686067964054038,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 28086
    },
    {
      "epoch": 6.6863060167827175,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 28087
    },
    {
      "epoch": 6.686544069511397,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 28088
    },
    {
      "epoch": 6.686782122240076,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 28089
    },
    {
      "epoch": 6.6870201749687554,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 28090
    },
    {
      "epoch": 6.687258227697435,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 28091
    },
    {
      "epoch": 6.687496280426115,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 28092
    },
    {
      "epoch": 6.687734333154793,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 28093
    },
    {
      "epoch": 6.687972385883473,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 28094
    },
    {
      "epoch": 6.688210438612153,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 28095
    },
    {
      "epoch": 6.688448491340832,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 28096
    },
    {
      "epoch": 6.688686544069512,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 28097
    },
    {
      "epoch": 6.6889245967981905,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 28098
    },
    {
      "epoch": 6.68916264952687,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 28099
    },
    {
      "epoch": 6.68940070225555,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 28100
    },
    {
      "epoch": 6.68940070225555,
      "eval_runtime": 0.514,
      "eval_samples_per_second": 1.945,
      "eval_steps_per_second": 1.945,
      "step": 28100
    },
    {
      "epoch": 6.689638754984229,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 28101
    },
    {
      "epoch": 6.689876807712908,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 28102
    },
    {
      "epoch": 6.690114860441588,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.368,
      "step": 28103
    },
    {
      "epoch": 6.690352913170267,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 28104
    },
    {
      "epoch": 6.690590965898947,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 28105
    },
    {
      "epoch": 6.6908290186276265,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 28106
    },
    {
      "epoch": 6.691067071356305,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 28107
    },
    {
      "epoch": 6.691305124084985,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 28108
    },
    {
      "epoch": 6.691543176813664,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 28109
    },
    {
      "epoch": 6.691781229542344,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 28110
    },
    {
      "epoch": 6.692019282271023,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8952,
      "step": 28111
    },
    {
      "epoch": 6.692257334999702,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 28112
    },
    {
      "epoch": 6.692495387728382,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 28113
    },
    {
      "epoch": 6.6927334404570615,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8053,
      "step": 28114
    },
    {
      "epoch": 6.692971493185741,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 28115
    },
    {
      "epoch": 6.69320954591442,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 28116
    },
    {
      "epoch": 6.6934475986430995,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 28117
    },
    {
      "epoch": 6.693685651371779,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 28118
    },
    {
      "epoch": 6.693923704100458,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 28119
    },
    {
      "epoch": 6.694161756829137,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 28120
    },
    {
      "epoch": 6.694399809557817,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 28121
    },
    {
      "epoch": 6.694637862286497,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3132,
      "step": 28122
    },
    {
      "epoch": 6.694875915015176,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 28123
    },
    {
      "epoch": 6.695113967743855,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 28124
    },
    {
      "epoch": 6.6953520204725345,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2934,
      "step": 28125
    },
    {
      "epoch": 6.695590073201214,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 28126
    },
    {
      "epoch": 6.695828125929894,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 28127
    },
    {
      "epoch": 6.6960661786585725,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 28128
    },
    {
      "epoch": 6.696304231387252,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 28129
    },
    {
      "epoch": 6.696542284115932,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 28130
    },
    {
      "epoch": 6.696780336844611,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 28131
    },
    {
      "epoch": 6.697018389573291,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 28132
    },
    {
      "epoch": 6.69725644230197,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 28133
    },
    {
      "epoch": 6.697494495030649,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 28134
    },
    {
      "epoch": 6.697732547759329,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 28135
    },
    {
      "epoch": 6.697970600488008,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 28136
    },
    {
      "epoch": 6.698208653216687,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 28137
    },
    {
      "epoch": 6.698446705945367,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 28138
    },
    {
      "epoch": 6.698684758674046,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 28139
    },
    {
      "epoch": 6.698922811402726,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 28140
    },
    {
      "epoch": 6.6991608641314055,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 28141
    },
    {
      "epoch": 6.699398916860084,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 28142
    },
    {
      "epoch": 6.699636969588764,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 28143
    },
    {
      "epoch": 6.6998750223174435,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 28144
    },
    {
      "epoch": 6.700113075046123,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 28145
    },
    {
      "epoch": 6.700351127774802,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2069,
      "step": 28146
    },
    {
      "epoch": 6.700589180503481,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 28147
    },
    {
      "epoch": 6.700827233232161,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7621,
      "step": 28148
    },
    {
      "epoch": 6.701065285960841,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 28149
    },
    {
      "epoch": 6.70130333868952,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 28150
    },
    {
      "epoch": 6.701541391418199,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 28151
    },
    {
      "epoch": 6.7017794441468785,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 28152
    },
    {
      "epoch": 6.702017496875558,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.5267,
      "step": 28153
    },
    {
      "epoch": 6.702255549604237,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 28154
    },
    {
      "epoch": 6.7024936023329165,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 28155
    },
    {
      "epoch": 6.702731655061596,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 28156
    },
    {
      "epoch": 6.702969707790276,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 28157
    },
    {
      "epoch": 6.703207760518955,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2746,
      "step": 28158
    },
    {
      "epoch": 6.703445813247634,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8245,
      "step": 28159
    },
    {
      "epoch": 6.703683865976314,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 28160
    },
    {
      "epoch": 6.703921918704993,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8852,
      "step": 28161
    },
    {
      "epoch": 6.704159971433673,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 28162
    },
    {
      "epoch": 6.7043980241623515,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 28163
    },
    {
      "epoch": 6.704636076891031,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3437,
      "step": 28164
    },
    {
      "epoch": 6.704874129619711,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 28165
    },
    {
      "epoch": 6.70511218234839,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 28166
    },
    {
      "epoch": 6.70535023507707,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 28167
    },
    {
      "epoch": 6.705588287805749,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 28168
    },
    {
      "epoch": 6.705826340534428,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 28169
    },
    {
      "epoch": 6.706064393263108,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2486,
      "step": 28170
    },
    {
      "epoch": 6.7063024459917875,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 28171
    },
    {
      "epoch": 6.706540498720466,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 28172
    },
    {
      "epoch": 6.706778551449146,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 28173
    },
    {
      "epoch": 6.707016604177825,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 28174
    },
    {
      "epoch": 6.707254656906505,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2972,
      "step": 28175
    },
    {
      "epoch": 6.707492709635185,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 28176
    },
    {
      "epoch": 6.707730762363863,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 28177
    },
    {
      "epoch": 6.707968815092543,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 28178
    },
    {
      "epoch": 6.7082068678212226,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 28179
    },
    {
      "epoch": 6.708444920549902,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3156,
      "step": 28180
    },
    {
      "epoch": 6.708682973278581,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 28181
    },
    {
      "epoch": 6.7089210260072605,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3244,
      "step": 28182
    },
    {
      "epoch": 6.70915907873594,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 28183
    },
    {
      "epoch": 6.70939713146462,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.288,
      "step": 28184
    },
    {
      "epoch": 6.709635184193299,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 28185
    },
    {
      "epoch": 6.709873236921978,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 28186
    },
    {
      "epoch": 6.710111289650658,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 28187
    },
    {
      "epoch": 6.710349342379337,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1812,
      "step": 28188
    },
    {
      "epoch": 6.710587395108016,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 28189
    },
    {
      "epoch": 6.7108254478366955,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 28190
    },
    {
      "epoch": 6.711063500565375,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 28191
    },
    {
      "epoch": 6.711301553294055,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1825,
      "step": 28192
    },
    {
      "epoch": 6.711539606022734,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 28193
    },
    {
      "epoch": 6.711777658751413,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 28194
    },
    {
      "epoch": 6.712015711480093,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 28195
    },
    {
      "epoch": 6.712253764208772,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 28196
    },
    {
      "epoch": 6.712491816937452,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 28197
    },
    {
      "epoch": 6.712729869666131,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 28198
    },
    {
      "epoch": 6.71296792239481,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 28199
    },
    {
      "epoch": 6.71320597512349,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 28200
    },
    {
      "epoch": 6.71320597512349,
      "eval_runtime": 0.5298,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 28200
    },
    {
      "epoch": 6.713444027852169,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 28201
    },
    {
      "epoch": 6.713682080580849,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 28202
    },
    {
      "epoch": 6.713920133309528,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 28203
    },
    {
      "epoch": 6.714158186038207,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 28204
    },
    {
      "epoch": 6.714396238766887,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 28205
    },
    {
      "epoch": 6.714634291495567,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 28206
    },
    {
      "epoch": 6.714872344224245,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 28207
    },
    {
      "epoch": 6.715110396952925,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 28208
    },
    {
      "epoch": 6.7153484496816045,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 28209
    },
    {
      "epoch": 6.715586502410284,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 28210
    },
    {
      "epoch": 6.715824555138964,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 28211
    },
    {
      "epoch": 6.716062607867642,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4037,
      "step": 28212
    },
    {
      "epoch": 6.716300660596322,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 28213
    },
    {
      "epoch": 6.716538713325002,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2263,
      "step": 28214
    },
    {
      "epoch": 6.716776766053681,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2391,
      "step": 28215
    },
    {
      "epoch": 6.71701481878236,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 28216
    },
    {
      "epoch": 6.71725287151104,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 28217
    },
    {
      "epoch": 6.717490924239719,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8687,
      "step": 28218
    },
    {
      "epoch": 6.717728976968399,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 28219
    },
    {
      "epoch": 6.717967029697078,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 28220
    },
    {
      "epoch": 6.718205082425757,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 28221
    },
    {
      "epoch": 6.718443135154437,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 28222
    },
    {
      "epoch": 6.718681187883116,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 28223
    },
    {
      "epoch": 6.718919240611795,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 28224
    },
    {
      "epoch": 6.719157293340475,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2038,
      "step": 28225
    },
    {
      "epoch": 6.719395346069154,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 28226
    },
    {
      "epoch": 6.719633398797834,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2259,
      "step": 28227
    },
    {
      "epoch": 6.719871451526513,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 28228
    },
    {
      "epoch": 6.720109504255192,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 28229
    },
    {
      "epoch": 6.720347556983872,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 28230
    },
    {
      "epoch": 6.720585609712551,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 28231
    },
    {
      "epoch": 6.720823662441231,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 28232
    },
    {
      "epoch": 6.72106171516991,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 28233
    },
    {
      "epoch": 6.721299767898589,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 28234
    },
    {
      "epoch": 6.721537820627269,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 28235
    },
    {
      "epoch": 6.7217758733559485,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 28236
    },
    {
      "epoch": 6.722013926084628,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.45,
      "step": 28237
    },
    {
      "epoch": 6.722251978813307,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 28238
    },
    {
      "epoch": 6.722490031541986,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 28239
    },
    {
      "epoch": 6.722728084270666,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 28240
    },
    {
      "epoch": 6.722966136999346,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 28241
    },
    {
      "epoch": 6.723204189728024,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 28242
    },
    {
      "epoch": 6.723442242456704,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 28243
    },
    {
      "epoch": 6.723680295185384,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 28244
    },
    {
      "epoch": 6.723918347914063,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 28245
    },
    {
      "epoch": 6.724156400642743,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 28246
    },
    {
      "epoch": 6.7243944533714215,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 28247
    },
    {
      "epoch": 6.724632506100101,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 28248
    },
    {
      "epoch": 6.724870558828781,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 28249
    },
    {
      "epoch": 6.72510861155746,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 28250
    },
    {
      "epoch": 6.725346664286139,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 28251
    },
    {
      "epoch": 6.725584717014819,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 28252
    },
    {
      "epoch": 6.725822769743498,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 28253
    },
    {
      "epoch": 6.726060822472178,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2523,
      "step": 28254
    },
    {
      "epoch": 6.7262988752008575,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 28255
    },
    {
      "epoch": 6.726536927929536,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 28256
    },
    {
      "epoch": 6.726774980658216,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 28257
    },
    {
      "epoch": 6.727013033386895,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 28258
    },
    {
      "epoch": 6.727251086115575,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 28259
    },
    {
      "epoch": 6.727489138844254,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8164,
      "step": 28260
    },
    {
      "epoch": 6.727727191572933,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.4302,
      "step": 28261
    },
    {
      "epoch": 6.727965244301613,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 28262
    },
    {
      "epoch": 6.7282032970302925,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7704,
      "step": 28263
    },
    {
      "epoch": 6.728441349758971,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 28264
    },
    {
      "epoch": 6.728679402487651,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 28265
    },
    {
      "epoch": 6.7289174552163304,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 28266
    },
    {
      "epoch": 6.72915550794501,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 28267
    },
    {
      "epoch": 6.729393560673689,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 28268
    },
    {
      "epoch": 6.729631613402368,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2811,
      "step": 28269
    },
    {
      "epoch": 6.729869666131048,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 28270
    },
    {
      "epoch": 6.730107718859728,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 28271
    },
    {
      "epoch": 6.730345771588407,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 28272
    },
    {
      "epoch": 6.730583824317086,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 28273
    },
    {
      "epoch": 6.7308218770457655,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 28274
    },
    {
      "epoch": 6.731059929774445,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 28275
    },
    {
      "epoch": 6.731297982503125,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 28276
    },
    {
      "epoch": 6.731536035231803,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1442,
      "step": 28277
    },
    {
      "epoch": 6.731774087960483,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 28278
    },
    {
      "epoch": 6.732012140689163,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 28279
    },
    {
      "epoch": 6.732250193417842,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2105,
      "step": 28280
    },
    {
      "epoch": 6.732488246146522,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 28281
    },
    {
      "epoch": 6.732726298875201,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 28282
    },
    {
      "epoch": 6.73296435160388,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 28283
    },
    {
      "epoch": 6.73320240433256,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 28284
    },
    {
      "epoch": 6.733440457061239,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 28285
    },
    {
      "epoch": 6.733678509789918,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 28286
    },
    {
      "epoch": 6.733916562518598,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 28287
    },
    {
      "epoch": 6.734154615247277,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9463,
      "step": 28288
    },
    {
      "epoch": 6.734392667975957,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2495,
      "step": 28289
    },
    {
      "epoch": 6.7346307207046365,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 28290
    },
    {
      "epoch": 6.734868773433315,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 28291
    },
    {
      "epoch": 6.735106826161995,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 28292
    },
    {
      "epoch": 6.7353448788906745,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7973,
      "step": 28293
    },
    {
      "epoch": 6.735582931619354,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 28294
    },
    {
      "epoch": 6.735820984348033,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 28295
    },
    {
      "epoch": 6.736059037076712,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 28296
    },
    {
      "epoch": 6.736297089805392,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 28297
    },
    {
      "epoch": 6.736535142534072,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 28298
    },
    {
      "epoch": 6.73677319526275,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 28299
    },
    {
      "epoch": 6.73701124799143,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 28300
    },
    {
      "epoch": 6.73701124799143,
      "eval_runtime": 0.5287,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 28300
    },
    {
      "epoch": 6.7372493007201095,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 28301
    },
    {
      "epoch": 6.737487353448789,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 28302
    },
    {
      "epoch": 6.737725406177468,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 28303
    },
    {
      "epoch": 6.7379634589061475,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 28304
    },
    {
      "epoch": 6.738201511634827,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 28305
    },
    {
      "epoch": 6.738439564363507,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 28306
    },
    {
      "epoch": 6.738677617092186,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 28307
    },
    {
      "epoch": 6.738915669820865,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3768,
      "step": 28308
    },
    {
      "epoch": 6.739153722549545,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 28309
    },
    {
      "epoch": 6.739391775278224,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.157,
      "step": 28310
    },
    {
      "epoch": 6.739629828006904,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 28311
    },
    {
      "epoch": 6.7398678807355825,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 28312
    },
    {
      "epoch": 6.740105933464262,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7852,
      "step": 28313
    },
    {
      "epoch": 6.740343986192942,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7919,
      "step": 28314
    },
    {
      "epoch": 6.740582038921621,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 28315
    },
    {
      "epoch": 6.740820091650301,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 28316
    },
    {
      "epoch": 6.74105814437898,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 28317
    },
    {
      "epoch": 6.741296197107659,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3353,
      "step": 28318
    },
    {
      "epoch": 6.741534249836339,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 28319
    },
    {
      "epoch": 6.7417723025650185,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 28320
    },
    {
      "epoch": 6.742010355293697,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2988,
      "step": 28321
    },
    {
      "epoch": 6.742248408022377,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 28322
    },
    {
      "epoch": 6.742486460751056,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 28323
    },
    {
      "epoch": 6.742724513479736,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 28324
    },
    {
      "epoch": 6.742962566208416,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3213,
      "step": 28325
    },
    {
      "epoch": 6.743200618937094,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 28326
    },
    {
      "epoch": 6.743438671665774,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 28327
    },
    {
      "epoch": 6.7436767243944535,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4267,
      "step": 28328
    },
    {
      "epoch": 6.743914777123133,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 28329
    },
    {
      "epoch": 6.744152829851812,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 28330
    },
    {
      "epoch": 6.7443908825804915,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 28331
    },
    {
      "epoch": 6.744628935309171,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3826,
      "step": 28332
    },
    {
      "epoch": 6.744866988037851,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 28333
    },
    {
      "epoch": 6.74510504076653,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 28334
    },
    {
      "epoch": 6.745343093495209,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 28335
    },
    {
      "epoch": 6.745581146223889,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.314,
      "step": 28336
    },
    {
      "epoch": 6.745819198952568,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 28337
    },
    {
      "epoch": 6.746057251681247,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 28338
    },
    {
      "epoch": 6.7462953044099265,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.4601,
      "step": 28339
    },
    {
      "epoch": 6.746533357138606,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 28340
    },
    {
      "epoch": 6.746771409867286,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 28341
    },
    {
      "epoch": 6.747009462595965,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 28342
    },
    {
      "epoch": 6.747247515324644,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 28343
    },
    {
      "epoch": 6.747485568053324,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 28344
    },
    {
      "epoch": 6.747723620782003,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 28345
    },
    {
      "epoch": 6.747961673510683,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 28346
    },
    {
      "epoch": 6.748199726239362,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 28347
    },
    {
      "epoch": 6.748437778968041,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 28348
    },
    {
      "epoch": 6.748675831696721,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 28349
    },
    {
      "epoch": 6.7489138844254,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8767,
      "step": 28350
    },
    {
      "epoch": 6.74915193715408,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 28351
    },
    {
      "epoch": 6.749389989882759,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 28352
    },
    {
      "epoch": 6.749628042611438,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 28353
    },
    {
      "epoch": 6.749866095340118,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 28354
    },
    {
      "epoch": 6.7501041480687975,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 28355
    },
    {
      "epoch": 6.750342200797476,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 28356
    },
    {
      "epoch": 6.750580253526156,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 28357
    },
    {
      "epoch": 6.7508183062548355,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 28358
    },
    {
      "epoch": 6.751056358983515,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 28359
    },
    {
      "epoch": 6.751294411712195,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 28360
    },
    {
      "epoch": 6.751532464440873,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 28361
    },
    {
      "epoch": 6.751770517169553,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 28362
    },
    {
      "epoch": 6.752008569898233,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 28363
    },
    {
      "epoch": 6.752246622626912,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 28364
    },
    {
      "epoch": 6.752484675355591,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 28365
    },
    {
      "epoch": 6.7527227280842705,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 28366
    },
    {
      "epoch": 6.75296078081295,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 28367
    },
    {
      "epoch": 6.75319883354163,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 28368
    },
    {
      "epoch": 6.753436886270309,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 28369
    },
    {
      "epoch": 6.753674938998988,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 28370
    },
    {
      "epoch": 6.753912991727668,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 28371
    },
    {
      "epoch": 6.754151044456347,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 28372
    },
    {
      "epoch": 6.754389097185026,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3168,
      "step": 28373
    },
    {
      "epoch": 6.754627149913706,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3147,
      "step": 28374
    },
    {
      "epoch": 6.754865202642385,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 28375
    },
    {
      "epoch": 6.755103255371065,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 28376
    },
    {
      "epoch": 6.755341308099744,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 28377
    },
    {
      "epoch": 6.755579360828423,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.319,
      "step": 28378
    },
    {
      "epoch": 6.755817413557103,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.935,
      "step": 28379
    },
    {
      "epoch": 6.756055466285782,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 28380
    },
    {
      "epoch": 6.756293519014462,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 28381
    },
    {
      "epoch": 6.756531571743141,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 28382
    },
    {
      "epoch": 6.75676962447182,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 28383
    },
    {
      "epoch": 6.7570076772005,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 28384
    },
    {
      "epoch": 6.7572457299291795,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 28385
    },
    {
      "epoch": 6.757483782657859,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2071,
      "step": 28386
    },
    {
      "epoch": 6.757721835386538,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 28387
    },
    {
      "epoch": 6.757959888115217,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 28388
    },
    {
      "epoch": 6.758197940843897,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 28389
    },
    {
      "epoch": 6.758435993572577,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4343,
      "step": 28390
    },
    {
      "epoch": 6.758674046301255,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 28391
    },
    {
      "epoch": 6.758912099029935,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 28392
    },
    {
      "epoch": 6.7591501517586146,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 28393
    },
    {
      "epoch": 6.759388204487294,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8692,
      "step": 28394
    },
    {
      "epoch": 6.759626257215974,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 28395
    },
    {
      "epoch": 6.7598643099446525,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 28396
    },
    {
      "epoch": 6.760102362673332,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.26,
      "step": 28397
    },
    {
      "epoch": 6.760340415402012,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 28398
    },
    {
      "epoch": 6.760578468130691,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 28399
    },
    {
      "epoch": 6.76081652085937,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 28400
    },
    {
      "epoch": 6.76081652085937,
      "eval_runtime": 0.5207,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 28400
    },
    {
      "epoch": 6.76105457358805,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 28401
    },
    {
      "epoch": 6.761292626316729,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 28402
    },
    {
      "epoch": 6.761530679045409,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8347,
      "step": 28403
    },
    {
      "epoch": 6.761768731774088,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.344,
      "step": 28404
    },
    {
      "epoch": 6.762006784502767,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 28405
    },
    {
      "epoch": 6.762244837231447,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.336,
      "step": 28406
    },
    {
      "epoch": 6.762482889960126,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 28407
    },
    {
      "epoch": 6.762720942688805,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 28408
    },
    {
      "epoch": 6.762958995417485,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3047,
      "step": 28409
    },
    {
      "epoch": 6.763197048146164,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 28410
    },
    {
      "epoch": 6.763435100874844,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 28411
    },
    {
      "epoch": 6.7636731536035235,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 28412
    },
    {
      "epoch": 6.763911206332202,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 28413
    },
    {
      "epoch": 6.764149259060882,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 28414
    },
    {
      "epoch": 6.764387311789561,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 28415
    },
    {
      "epoch": 6.764625364518241,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 28416
    },
    {
      "epoch": 6.76486341724692,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 28417
    },
    {
      "epoch": 6.765101469975599,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 28418
    },
    {
      "epoch": 6.765339522704279,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 28419
    },
    {
      "epoch": 6.765577575432959,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 28420
    },
    {
      "epoch": 6.765815628161638,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 28421
    },
    {
      "epoch": 6.766053680890317,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 28422
    },
    {
      "epoch": 6.7662917336189965,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8497,
      "step": 28423
    },
    {
      "epoch": 6.766529786347676,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 28424
    },
    {
      "epoch": 6.766767839076356,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2599,
      "step": 28425
    },
    {
      "epoch": 6.767005891805034,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 28426
    },
    {
      "epoch": 6.767243944533714,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9118,
      "step": 28427
    },
    {
      "epoch": 6.767481997262394,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 28428
    },
    {
      "epoch": 6.767720049991073,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1646,
      "step": 28429
    },
    {
      "epoch": 6.767958102719753,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 28430
    },
    {
      "epoch": 6.768196155448432,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 28431
    },
    {
      "epoch": 6.768434208177111,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 28432
    },
    {
      "epoch": 6.768672260905791,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 28433
    },
    {
      "epoch": 6.76891031363447,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 28434
    },
    {
      "epoch": 6.769148366363149,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 28435
    },
    {
      "epoch": 6.769386419091829,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3768,
      "step": 28436
    },
    {
      "epoch": 6.769624471820508,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 28437
    },
    {
      "epoch": 6.769862524549188,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.4264,
      "step": 28438
    },
    {
      "epoch": 6.7701005772778675,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 28439
    },
    {
      "epoch": 6.770338630006546,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 28440
    },
    {
      "epoch": 6.770576682735226,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 28441
    },
    {
      "epoch": 6.770814735463905,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 28442
    },
    {
      "epoch": 6.771052788192584,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 28443
    },
    {
      "epoch": 6.771290840921264,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 28444
    },
    {
      "epoch": 6.771528893649943,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1627,
      "step": 28445
    },
    {
      "epoch": 6.771766946378623,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 28446
    },
    {
      "epoch": 6.772004999107303,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2545,
      "step": 28447
    },
    {
      "epoch": 6.772243051835981,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 28448
    },
    {
      "epoch": 6.772481104564661,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 28449
    },
    {
      "epoch": 6.7727191572933405,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 28450
    },
    {
      "epoch": 6.77295721002202,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 28451
    },
    {
      "epoch": 6.773195262750699,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 28452
    },
    {
      "epoch": 6.773433315479378,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 28453
    },
    {
      "epoch": 6.773671368208058,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 28454
    },
    {
      "epoch": 6.773909420936738,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8594,
      "step": 28455
    },
    {
      "epoch": 6.774147473665417,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 28456
    },
    {
      "epoch": 6.774385526394096,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3529,
      "step": 28457
    },
    {
      "epoch": 6.774623579122776,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 28458
    },
    {
      "epoch": 6.774861631851455,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4146,
      "step": 28459
    },
    {
      "epoch": 6.775099684580135,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 28460
    },
    {
      "epoch": 6.7753377373088135,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.281,
      "step": 28461
    },
    {
      "epoch": 6.775575790037493,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 28462
    },
    {
      "epoch": 6.775813842766173,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 28463
    },
    {
      "epoch": 6.776051895494852,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 28464
    },
    {
      "epoch": 6.776289948223532,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 28465
    },
    {
      "epoch": 6.776528000952211,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 28466
    },
    {
      "epoch": 6.77676605368089,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9151,
      "step": 28467
    },
    {
      "epoch": 6.77700410640957,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 28468
    },
    {
      "epoch": 6.7772421591382495,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 28469
    },
    {
      "epoch": 6.777480211866928,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9063,
      "step": 28470
    },
    {
      "epoch": 6.777718264595608,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 28471
    },
    {
      "epoch": 6.777956317324287,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 28472
    },
    {
      "epoch": 6.778194370052967,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 28473
    },
    {
      "epoch": 6.778432422781647,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 28474
    },
    {
      "epoch": 6.778670475510325,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 28475
    },
    {
      "epoch": 6.778908528239005,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8785,
      "step": 28476
    },
    {
      "epoch": 6.7791465809676845,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 28477
    },
    {
      "epoch": 6.779384633696364,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 28478
    },
    {
      "epoch": 6.779622686425043,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 28479
    },
    {
      "epoch": 6.7798607391537224,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 28480
    },
    {
      "epoch": 6.780098791882402,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 28481
    },
    {
      "epoch": 6.780336844611082,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7235,
      "step": 28482
    },
    {
      "epoch": 6.78057489733976,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 28483
    },
    {
      "epoch": 6.78081295006844,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2287,
      "step": 28484
    },
    {
      "epoch": 6.78105100279712,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 28485
    },
    {
      "epoch": 6.781289055525799,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 28486
    },
    {
      "epoch": 6.781527108254478,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 28487
    },
    {
      "epoch": 6.7817651609831575,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 28488
    },
    {
      "epoch": 6.782003213711837,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 28489
    },
    {
      "epoch": 6.782241266440517,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 28490
    },
    {
      "epoch": 6.782479319169196,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 28491
    },
    {
      "epoch": 6.782717371897875,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 28492
    },
    {
      "epoch": 6.782955424626555,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 28493
    },
    {
      "epoch": 6.783193477355234,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 28494
    },
    {
      "epoch": 6.783431530083914,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 28495
    },
    {
      "epoch": 6.783669582812593,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.795,
      "step": 28496
    },
    {
      "epoch": 6.783907635541272,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1757,
      "step": 28497
    },
    {
      "epoch": 6.784145688269952,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 28498
    },
    {
      "epoch": 6.784383740998631,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 28499
    },
    {
      "epoch": 6.784621793727311,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 28500
    },
    {
      "epoch": 6.784621793727311,
      "eval_runtime": 0.5395,
      "eval_samples_per_second": 1.854,
      "eval_steps_per_second": 1.854,
      "step": 28500
    },
    {
      "epoch": 6.78485984645599,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 28501
    },
    {
      "epoch": 6.785097899184669,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 28502
    },
    {
      "epoch": 6.785335951913349,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2234,
      "step": 28503
    },
    {
      "epoch": 6.7855740046420285,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 28504
    },
    {
      "epoch": 6.785812057370707,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 28505
    },
    {
      "epoch": 6.786050110099387,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 28506
    },
    {
      "epoch": 6.7862881628280665,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 28507
    },
    {
      "epoch": 6.786526215556746,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 28508
    },
    {
      "epoch": 6.786764268285426,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0772,
      "step": 28509
    },
    {
      "epoch": 6.787002321014104,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 28510
    },
    {
      "epoch": 6.787240373742784,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 28511
    },
    {
      "epoch": 6.787478426471464,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 28512
    },
    {
      "epoch": 6.787716479200143,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 28513
    },
    {
      "epoch": 6.787954531928822,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 28514
    },
    {
      "epoch": 6.7881925846575015,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 28515
    },
    {
      "epoch": 6.788430637386181,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2629,
      "step": 28516
    },
    {
      "epoch": 6.788668690114861,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 28517
    },
    {
      "epoch": 6.78890674284354,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 28518
    },
    {
      "epoch": 6.789144795572219,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 28519
    },
    {
      "epoch": 6.789382848300899,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 28520
    },
    {
      "epoch": 6.789620901029578,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 28521
    },
    {
      "epoch": 6.789858953758257,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 28522
    },
    {
      "epoch": 6.790097006486937,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 28523
    },
    {
      "epoch": 6.790335059215616,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 28524
    },
    {
      "epoch": 6.790573111944296,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 28525
    },
    {
      "epoch": 6.790811164672975,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 28526
    },
    {
      "epoch": 6.791049217401654,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 28527
    },
    {
      "epoch": 6.791287270130334,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 28528
    },
    {
      "epoch": 6.791525322859013,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 28529
    },
    {
      "epoch": 6.791763375587693,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 28530
    },
    {
      "epoch": 6.792001428316372,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 28531
    },
    {
      "epoch": 6.792239481045051,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3335,
      "step": 28532
    },
    {
      "epoch": 6.792477533773731,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 28533
    },
    {
      "epoch": 6.7927155865024105,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8945,
      "step": 28534
    },
    {
      "epoch": 6.79295363923109,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 28535
    },
    {
      "epoch": 6.793191691959769,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 28536
    },
    {
      "epoch": 6.793429744688448,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 28537
    },
    {
      "epoch": 6.793667797417128,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 28538
    },
    {
      "epoch": 6.793905850145808,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 28539
    },
    {
      "epoch": 6.794143902874486,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 28540
    },
    {
      "epoch": 6.794381955603166,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 28541
    },
    {
      "epoch": 6.7946200083318455,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 28542
    },
    {
      "epoch": 6.794858061060525,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 28543
    },
    {
      "epoch": 6.795096113789205,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.3008,
      "step": 28544
    },
    {
      "epoch": 6.7953341665178835,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 28545
    },
    {
      "epoch": 6.795572219246563,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8062,
      "step": 28546
    },
    {
      "epoch": 6.795810271975243,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 28547
    },
    {
      "epoch": 6.796048324703922,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 28548
    },
    {
      "epoch": 6.796286377432601,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 28549
    },
    {
      "epoch": 6.796524430161281,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2888,
      "step": 28550
    },
    {
      "epoch": 6.79676248288996,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 28551
    },
    {
      "epoch": 6.79700053561864,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 28552
    },
    {
      "epoch": 6.797238588347319,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 28553
    },
    {
      "epoch": 6.797476641075998,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 28554
    },
    {
      "epoch": 6.797714693804678,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 28555
    },
    {
      "epoch": 6.797952746533357,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 28556
    },
    {
      "epoch": 6.798190799262036,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3187,
      "step": 28557
    },
    {
      "epoch": 6.798428851990716,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 28558
    },
    {
      "epoch": 6.798666904719395,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 28559
    },
    {
      "epoch": 6.798904957448075,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 28560
    },
    {
      "epoch": 6.7991430101767545,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2974,
      "step": 28561
    },
    {
      "epoch": 6.799381062905433,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 28562
    },
    {
      "epoch": 6.799619115634113,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 28563
    },
    {
      "epoch": 6.799857168362792,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4139,
      "step": 28564
    },
    {
      "epoch": 6.800095221091472,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 28565
    },
    {
      "epoch": 6.800333273820151,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5095,
      "step": 28566
    },
    {
      "epoch": 6.80057132654883,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 28567
    },
    {
      "epoch": 6.80080937927751,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 28568
    },
    {
      "epoch": 6.8010474320061896,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9285,
      "step": 28569
    },
    {
      "epoch": 6.801285484734869,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 28570
    },
    {
      "epoch": 6.801523537463548,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 28571
    },
    {
      "epoch": 6.8017615901922275,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 28572
    },
    {
      "epoch": 6.801999642920907,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 28573
    },
    {
      "epoch": 6.802237695649587,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 28574
    },
    {
      "epoch": 6.802475748378265,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3149,
      "step": 28575
    },
    {
      "epoch": 6.802713801106945,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 28576
    },
    {
      "epoch": 6.802951853835625,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 28577
    },
    {
      "epoch": 6.803189906564304,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 28578
    },
    {
      "epoch": 6.803427959292984,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0176,
      "step": 28579
    },
    {
      "epoch": 6.8036660120216625,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 28580
    },
    {
      "epoch": 6.803904064750342,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 28581
    },
    {
      "epoch": 6.804142117479022,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 28582
    },
    {
      "epoch": 6.804380170207701,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 28583
    },
    {
      "epoch": 6.80461822293638,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 28584
    },
    {
      "epoch": 6.80485627566506,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.138,
      "step": 28585
    },
    {
      "epoch": 6.805094328393739,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 28586
    },
    {
      "epoch": 6.805332381122419,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 28587
    },
    {
      "epoch": 6.8055704338510985,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7968,
      "step": 28588
    },
    {
      "epoch": 6.805808486579777,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 28589
    },
    {
      "epoch": 6.806046539308457,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7227,
      "step": 28590
    },
    {
      "epoch": 6.806284592037136,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 28591
    },
    {
      "epoch": 6.806522644765815,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 28592
    },
    {
      "epoch": 6.806760697494495,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 28593
    },
    {
      "epoch": 6.806998750223174,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 28594
    },
    {
      "epoch": 6.807236802951854,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 28595
    },
    {
      "epoch": 6.807474855680534,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 28596
    },
    {
      "epoch": 6.807712908409212,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 28597
    },
    {
      "epoch": 6.807950961137892,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 28598
    },
    {
      "epoch": 6.8081890138665715,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 28599
    },
    {
      "epoch": 6.808427066595251,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 28600
    },
    {
      "epoch": 6.808427066595251,
      "eval_runtime": 0.5282,
      "eval_samples_per_second": 1.893,
      "eval_steps_per_second": 1.893,
      "step": 28600
    },
    {
      "epoch": 6.80866511932393,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 28601
    },
    {
      "epoch": 6.808903172052609,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 28602
    },
    {
      "epoch": 6.809141224781289,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 28603
    },
    {
      "epoch": 6.809379277509969,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 28604
    },
    {
      "epoch": 6.809617330238648,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4639,
      "step": 28605
    },
    {
      "epoch": 6.809855382967327,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 28606
    },
    {
      "epoch": 6.810093435696007,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 28607
    },
    {
      "epoch": 6.810331488424686,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 28608
    },
    {
      "epoch": 6.810569541153366,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 28609
    },
    {
      "epoch": 6.8108075938820445,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 28610
    },
    {
      "epoch": 6.811045646610724,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 28611
    },
    {
      "epoch": 6.811283699339404,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 28612
    },
    {
      "epoch": 6.811521752068083,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 28613
    },
    {
      "epoch": 6.811759804796763,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 28614
    },
    {
      "epoch": 6.811997857525442,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 28615
    },
    {
      "epoch": 6.812235910254121,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 28616
    },
    {
      "epoch": 6.812473962982801,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 28617
    },
    {
      "epoch": 6.81271201571148,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 28618
    },
    {
      "epoch": 6.812950068440159,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 28619
    },
    {
      "epoch": 6.813188121168839,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9846,
      "step": 28620
    },
    {
      "epoch": 6.813426173897518,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 28621
    },
    {
      "epoch": 6.813664226626198,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2805,
      "step": 28622
    },
    {
      "epoch": 6.813902279354878,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 28623
    },
    {
      "epoch": 6.814140332083556,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 28624
    },
    {
      "epoch": 6.814378384812236,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.27,
      "step": 28625
    },
    {
      "epoch": 6.8146164375409155,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 28626
    },
    {
      "epoch": 6.814854490269594,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 28627
    },
    {
      "epoch": 6.815092542998274,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 28628
    },
    {
      "epoch": 6.815330595726953,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 28629
    },
    {
      "epoch": 6.815568648455633,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7783,
      "step": 28630
    },
    {
      "epoch": 6.815806701184313,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 28631
    },
    {
      "epoch": 6.816044753912991,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 28632
    },
    {
      "epoch": 6.816282806641671,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 28633
    },
    {
      "epoch": 6.816520859370351,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2119,
      "step": 28634
    },
    {
      "epoch": 6.81675891209903,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8181,
      "step": 28635
    },
    {
      "epoch": 6.816996964827709,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 28636
    },
    {
      "epoch": 6.8172350175563885,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 28637
    },
    {
      "epoch": 6.817473070285068,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 28638
    },
    {
      "epoch": 6.817711123013748,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 28639
    },
    {
      "epoch": 6.817949175742427,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3619,
      "step": 28640
    },
    {
      "epoch": 6.818187228471106,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 28641
    },
    {
      "epoch": 6.818425281199786,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8135,
      "step": 28642
    },
    {
      "epoch": 6.818663333928465,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 28643
    },
    {
      "epoch": 6.818901386657145,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 28644
    },
    {
      "epoch": 6.819139439385824,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4431,
      "step": 28645
    },
    {
      "epoch": 6.819377492114503,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.4051,
      "step": 28646
    },
    {
      "epoch": 6.819615544843183,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 28647
    },
    {
      "epoch": 6.819853597571862,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 28648
    },
    {
      "epoch": 6.820091650300542,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 28649
    },
    {
      "epoch": 6.820329703029221,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 28650
    },
    {
      "epoch": 6.8205677557579,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 28651
    },
    {
      "epoch": 6.82080580848658,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2252,
      "step": 28652
    },
    {
      "epoch": 6.8210438612152595,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 28653
    },
    {
      "epoch": 6.821281913943938,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 28654
    },
    {
      "epoch": 6.821519966672618,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 28655
    },
    {
      "epoch": 6.8217580194012974,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 28656
    },
    {
      "epoch": 6.821996072129977,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8491,
      "step": 28657
    },
    {
      "epoch": 6.822234124858657,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 28658
    },
    {
      "epoch": 6.822472177587335,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 28659
    },
    {
      "epoch": 6.822710230316015,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 28660
    },
    {
      "epoch": 6.822948283044695,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 28661
    },
    {
      "epoch": 6.823186335773374,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 28662
    },
    {
      "epoch": 6.823424388502053,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 28663
    },
    {
      "epoch": 6.8236624412307325,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 28664
    },
    {
      "epoch": 6.823900493959412,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9137,
      "step": 28665
    },
    {
      "epoch": 6.824138546688092,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 28666
    },
    {
      "epoch": 6.82437659941677,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 28667
    },
    {
      "epoch": 6.82461465214545,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 28668
    },
    {
      "epoch": 6.82485270487413,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 28669
    },
    {
      "epoch": 6.825090757602809,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 28670
    },
    {
      "epoch": 6.825328810331488,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 28671
    },
    {
      "epoch": 6.825566863060168,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 28672
    },
    {
      "epoch": 6.825804915788847,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 28673
    },
    {
      "epoch": 6.826042968517527,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 28674
    },
    {
      "epoch": 6.826281021246206,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 28675
    },
    {
      "epoch": 6.826519073974885,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3481,
      "step": 28676
    },
    {
      "epoch": 6.826757126703565,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 28677
    },
    {
      "epoch": 6.826995179432244,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 28678
    },
    {
      "epoch": 6.827233232160924,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 28679
    },
    {
      "epoch": 6.827471284889603,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 28680
    },
    {
      "epoch": 6.827709337618282,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 28681
    },
    {
      "epoch": 6.827947390346962,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2652,
      "step": 28682
    },
    {
      "epoch": 6.8281854430756415,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 28683
    },
    {
      "epoch": 6.828423495804321,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 28684
    },
    {
      "epoch": 6.828661548533,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.263,
      "step": 28685
    },
    {
      "epoch": 6.828899601261679,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 28686
    },
    {
      "epoch": 6.829137653990359,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2931,
      "step": 28687
    },
    {
      "epoch": 6.829375706719039,
      "grad_norm": 7.34375,
      "learning_rate": 3e-05,
      "loss": 2.4557,
      "step": 28688
    },
    {
      "epoch": 6.829613759447717,
      "grad_norm": 6.625,
      "learning_rate": 3e-05,
      "loss": 2.3155,
      "step": 28689
    },
    {
      "epoch": 6.829851812176397,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 28690
    },
    {
      "epoch": 6.8300898649050765,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 28691
    },
    {
      "epoch": 6.830327917633756,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 28692
    },
    {
      "epoch": 6.830565970362436,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.4726,
      "step": 28693
    },
    {
      "epoch": 6.8308040230911145,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 28694
    },
    {
      "epoch": 6.831042075819794,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 28695
    },
    {
      "epoch": 6.831280128548474,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2745,
      "step": 28696
    },
    {
      "epoch": 6.831518181277153,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 28697
    },
    {
      "epoch": 6.831756234005832,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 28698
    },
    {
      "epoch": 6.831994286734512,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 28699
    },
    {
      "epoch": 6.832232339463191,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 28700
    },
    {
      "epoch": 6.832232339463191,
      "eval_runtime": 0.5196,
      "eval_samples_per_second": 1.925,
      "eval_steps_per_second": 1.925,
      "step": 28700
    },
    {
      "epoch": 6.832470392191871,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 28701
    },
    {
      "epoch": 6.8327084449205495,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 28702
    },
    {
      "epoch": 6.832946497649229,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1516,
      "step": 28703
    },
    {
      "epoch": 6.833184550377909,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 28704
    },
    {
      "epoch": 6.833422603106588,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 28705
    },
    {
      "epoch": 6.833660655835267,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 28706
    },
    {
      "epoch": 6.833898708563947,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9725,
      "step": 28707
    },
    {
      "epoch": 6.834136761292626,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 28708
    },
    {
      "epoch": 6.834374814021306,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3156,
      "step": 28709
    },
    {
      "epoch": 6.8346128667499855,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3671,
      "step": 28710
    },
    {
      "epoch": 6.834850919478664,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 28711
    },
    {
      "epoch": 6.835088972207344,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2554,
      "step": 28712
    },
    {
      "epoch": 6.835327024936023,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 28713
    },
    {
      "epoch": 6.835565077664703,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 28714
    },
    {
      "epoch": 6.835803130393382,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2999,
      "step": 28715
    },
    {
      "epoch": 6.836041183122061,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 28716
    },
    {
      "epoch": 6.836279235850741,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6502,
      "step": 28717
    },
    {
      "epoch": 6.8365172885794205,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 28718
    },
    {
      "epoch": 6.8367553413081,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 28719
    },
    {
      "epoch": 6.836993394036779,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 28720
    },
    {
      "epoch": 6.8372314467654585,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 28721
    },
    {
      "epoch": 6.837469499494138,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 28722
    },
    {
      "epoch": 6.837707552222818,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 28723
    },
    {
      "epoch": 6.837945604951496,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3115,
      "step": 28724
    },
    {
      "epoch": 6.838183657680176,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9574,
      "step": 28725
    },
    {
      "epoch": 6.838421710408856,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3427,
      "step": 28726
    },
    {
      "epoch": 6.838659763137535,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 28727
    },
    {
      "epoch": 6.838897815866215,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 28728
    },
    {
      "epoch": 6.8391358685948935,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 28729
    },
    {
      "epoch": 6.839373921323573,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 28730
    },
    {
      "epoch": 6.839611974052253,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 28731
    },
    {
      "epoch": 6.839850026780932,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 28732
    },
    {
      "epoch": 6.840088079509611,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 28733
    },
    {
      "epoch": 6.840326132238291,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 28734
    },
    {
      "epoch": 6.84056418496697,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 28735
    },
    {
      "epoch": 6.84080223769565,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 28736
    },
    {
      "epoch": 6.8410402904243295,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 28737
    },
    {
      "epoch": 6.841278343153008,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3216,
      "step": 28738
    },
    {
      "epoch": 6.841516395881688,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 28739
    },
    {
      "epoch": 6.841754448610367,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 28740
    },
    {
      "epoch": 6.841992501339046,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 28741
    },
    {
      "epoch": 6.842230554067726,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4503,
      "step": 28742
    },
    {
      "epoch": 6.842468606796405,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 28743
    },
    {
      "epoch": 6.842706659525085,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 28744
    },
    {
      "epoch": 6.8429447122537645,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 28745
    },
    {
      "epoch": 6.843182764982443,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 28746
    },
    {
      "epoch": 6.843420817711123,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1831,
      "step": 28747
    },
    {
      "epoch": 6.8436588704398025,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 28748
    },
    {
      "epoch": 6.843896923168482,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 28749
    },
    {
      "epoch": 6.844134975897161,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 28750
    },
    {
      "epoch": 6.84437302862584,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 28751
    },
    {
      "epoch": 6.84461108135452,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 28752
    },
    {
      "epoch": 6.8448491340832,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 28753
    },
    {
      "epoch": 6.845087186811879,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2868,
      "step": 28754
    },
    {
      "epoch": 6.845325239540558,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 28755
    },
    {
      "epoch": 6.8455632922692375,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 28756
    },
    {
      "epoch": 6.845801344997917,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8619,
      "step": 28757
    },
    {
      "epoch": 6.846039397726597,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 28758
    },
    {
      "epoch": 6.8462774504552755,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 28759
    },
    {
      "epoch": 6.846515503183955,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 28760
    },
    {
      "epoch": 6.846753555912635,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 28761
    },
    {
      "epoch": 6.846991608641314,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 28762
    },
    {
      "epoch": 6.847229661369994,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 28763
    },
    {
      "epoch": 6.847467714098673,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 28764
    },
    {
      "epoch": 6.847705766827352,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 28765
    },
    {
      "epoch": 6.847943819556032,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3101,
      "step": 28766
    },
    {
      "epoch": 6.848181872284711,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1299,
      "step": 28767
    },
    {
      "epoch": 6.84841992501339,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0913,
      "step": 28768
    },
    {
      "epoch": 6.84865797774207,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8106,
      "step": 28769
    },
    {
      "epoch": 6.848896030470749,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 28770
    },
    {
      "epoch": 6.849134083199429,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9198,
      "step": 28771
    },
    {
      "epoch": 6.849372135928109,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 28772
    },
    {
      "epoch": 6.849610188656787,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 28773
    },
    {
      "epoch": 6.849848241385467,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 28774
    },
    {
      "epoch": 6.8500862941141465,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 28775
    },
    {
      "epoch": 6.850324346842825,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 28776
    },
    {
      "epoch": 6.850562399571505,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 28777
    },
    {
      "epoch": 6.850800452300184,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 28778
    },
    {
      "epoch": 6.851038505028864,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8753,
      "step": 28779
    },
    {
      "epoch": 6.851276557757544,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 28780
    },
    {
      "epoch": 6.851514610486222,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 28781
    },
    {
      "epoch": 6.851752663214902,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 28782
    },
    {
      "epoch": 6.8519907159435816,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8706,
      "step": 28783
    },
    {
      "epoch": 6.852228768672261,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 28784
    },
    {
      "epoch": 6.85246682140094,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2942,
      "step": 28785
    },
    {
      "epoch": 6.8527048741296195,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7419,
      "step": 28786
    },
    {
      "epoch": 6.852942926858299,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 28787
    },
    {
      "epoch": 6.853180979586979,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 28788
    },
    {
      "epoch": 6.853419032315658,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 28789
    },
    {
      "epoch": 6.853657085044337,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 28790
    },
    {
      "epoch": 6.853895137773017,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 28791
    },
    {
      "epoch": 6.854133190501696,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.97,
      "step": 28792
    },
    {
      "epoch": 6.854371243230376,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 28793
    },
    {
      "epoch": 6.8546092959590545,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3672,
      "step": 28794
    },
    {
      "epoch": 6.854847348687734,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8122,
      "step": 28795
    },
    {
      "epoch": 6.855085401416414,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 28796
    },
    {
      "epoch": 6.855323454145093,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4541,
      "step": 28797
    },
    {
      "epoch": 6.855561506873773,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9979,
      "step": 28798
    },
    {
      "epoch": 6.855799559602452,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 28799
    },
    {
      "epoch": 6.856037612331131,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 28800
    },
    {
      "epoch": 6.856037612331131,
      "eval_runtime": 0.5169,
      "eval_samples_per_second": 1.935,
      "eval_steps_per_second": 1.935,
      "step": 28800
    },
    {
      "epoch": 6.856275665059811,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 28801
    },
    {
      "epoch": 6.8565137177884905,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 28802
    },
    {
      "epoch": 6.856751770517169,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 28803
    },
    {
      "epoch": 6.856989823245849,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 28804
    },
    {
      "epoch": 6.857227875974528,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 28805
    },
    {
      "epoch": 6.857465928703208,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 28806
    },
    {
      "epoch": 6.857703981431888,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 28807
    },
    {
      "epoch": 6.857942034160566,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 28808
    },
    {
      "epoch": 6.858180086889246,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2224,
      "step": 28809
    },
    {
      "epoch": 6.858418139617926,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 28810
    },
    {
      "epoch": 6.858656192346604,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3277,
      "step": 28811
    },
    {
      "epoch": 6.858894245075284,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 28812
    },
    {
      "epoch": 6.8591322978039635,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 28813
    },
    {
      "epoch": 6.859370350532643,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 28814
    },
    {
      "epoch": 6.859608403261323,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 28815
    },
    {
      "epoch": 6.859846455990001,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 28816
    },
    {
      "epoch": 6.860084508718681,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 28817
    },
    {
      "epoch": 6.860322561447361,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 28818
    },
    {
      "epoch": 6.86056061417604,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5515,
      "step": 28819
    },
    {
      "epoch": 6.860798666904719,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 28820
    },
    {
      "epoch": 6.861036719633399,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8223,
      "step": 28821
    },
    {
      "epoch": 6.861274772362078,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2742,
      "step": 28822
    },
    {
      "epoch": 6.861512825090758,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 28823
    },
    {
      "epoch": 6.861750877819437,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8527,
      "step": 28824
    },
    {
      "epoch": 6.861988930548116,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 28825
    },
    {
      "epoch": 6.862226983276796,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 28826
    },
    {
      "epoch": 6.862465036005475,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 28827
    },
    {
      "epoch": 6.862703088734155,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 28828
    },
    {
      "epoch": 6.862941141462834,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 28829
    },
    {
      "epoch": 6.863179194191513,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 28830
    },
    {
      "epoch": 6.863417246920193,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2416,
      "step": 28831
    },
    {
      "epoch": 6.863655299648872,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9532,
      "step": 28832
    },
    {
      "epoch": 6.863893352377552,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 28833
    },
    {
      "epoch": 6.864131405106231,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 28834
    },
    {
      "epoch": 6.86436945783491,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3241,
      "step": 28835
    },
    {
      "epoch": 6.86460751056359,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 28836
    },
    {
      "epoch": 6.86484556329227,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 28837
    },
    {
      "epoch": 6.865083616020948,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 28838
    },
    {
      "epoch": 6.865321668749628,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 28839
    },
    {
      "epoch": 6.8655597214783075,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 28840
    },
    {
      "epoch": 6.865797774206987,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 28841
    },
    {
      "epoch": 6.866035826935667,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 28842
    },
    {
      "epoch": 6.866273879664345,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2984,
      "step": 28843
    },
    {
      "epoch": 6.866511932393025,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 28844
    },
    {
      "epoch": 6.866749985121705,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 28845
    },
    {
      "epoch": 6.866988037850383,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4377,
      "step": 28846
    },
    {
      "epoch": 6.867226090579063,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.047,
      "step": 28847
    },
    {
      "epoch": 6.867464143307743,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4469,
      "step": 28848
    },
    {
      "epoch": 6.867702196036422,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 28849
    },
    {
      "epoch": 6.867940248765102,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 28850
    },
    {
      "epoch": 6.8681783014937805,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3806,
      "step": 28851
    },
    {
      "epoch": 6.86841635422246,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 28852
    },
    {
      "epoch": 6.86865440695114,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6539,
      "step": 28853
    },
    {
      "epoch": 6.868892459679819,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 28854
    },
    {
      "epoch": 6.869130512408498,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 28855
    },
    {
      "epoch": 6.869368565137178,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 28856
    },
    {
      "epoch": 6.869606617865857,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 28857
    },
    {
      "epoch": 6.869844670594537,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 28858
    },
    {
      "epoch": 6.8700827233232165,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 28859
    },
    {
      "epoch": 6.870320776051895,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3386,
      "step": 28860
    },
    {
      "epoch": 6.870558828780575,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 28861
    },
    {
      "epoch": 6.870796881509254,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 28862
    },
    {
      "epoch": 6.871034934237934,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 28863
    },
    {
      "epoch": 6.871272986966613,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 28864
    },
    {
      "epoch": 6.871511039695292,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 28865
    },
    {
      "epoch": 6.871749092423972,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 28866
    },
    {
      "epoch": 6.8719871451526515,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 28867
    },
    {
      "epoch": 6.872225197881331,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 28868
    },
    {
      "epoch": 6.87246325061001,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 28869
    },
    {
      "epoch": 6.8727013033386894,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 28870
    },
    {
      "epoch": 6.872939356067369,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 28871
    },
    {
      "epoch": 6.873177408796049,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 28872
    },
    {
      "epoch": 6.873415461524727,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 28873
    },
    {
      "epoch": 6.873653514253407,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 28874
    },
    {
      "epoch": 6.873891566982087,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 28875
    },
    {
      "epoch": 6.874129619710766,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 28876
    },
    {
      "epoch": 6.874367672439446,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 28877
    },
    {
      "epoch": 6.8746057251681245,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 28878
    },
    {
      "epoch": 6.874843777896804,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 28879
    },
    {
      "epoch": 6.875081830625484,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 28880
    },
    {
      "epoch": 6.875319883354163,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 28881
    },
    {
      "epoch": 6.875557936082842,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 28882
    },
    {
      "epoch": 6.875795988811522,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 28883
    },
    {
      "epoch": 6.876034041540201,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 28884
    },
    {
      "epoch": 6.876272094268881,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 28885
    },
    {
      "epoch": 6.87651014699756,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 28886
    },
    {
      "epoch": 6.876748199726239,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 28887
    },
    {
      "epoch": 6.876986252454919,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 28888
    },
    {
      "epoch": 6.877224305183598,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 28889
    },
    {
      "epoch": 6.877462357912277,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 28890
    },
    {
      "epoch": 6.877700410640957,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.31,
      "step": 28891
    },
    {
      "epoch": 6.877938463369636,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 28892
    },
    {
      "epoch": 6.878176516098316,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 28893
    },
    {
      "epoch": 6.8784145688269955,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 28894
    },
    {
      "epoch": 6.878652621555674,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 28895
    },
    {
      "epoch": 6.878890674284354,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 28896
    },
    {
      "epoch": 6.8791287270130335,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.4314,
      "step": 28897
    },
    {
      "epoch": 6.879366779741713,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 28898
    },
    {
      "epoch": 6.879604832470392,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1297,
      "step": 28899
    },
    {
      "epoch": 6.879842885199071,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 28900
    },
    {
      "epoch": 6.879842885199071,
      "eval_runtime": 0.5316,
      "eval_samples_per_second": 1.881,
      "eval_steps_per_second": 1.881,
      "step": 28900
    },
    {
      "epoch": 6.880080937927751,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2312,
      "step": 28901
    },
    {
      "epoch": 6.880318990656431,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 28902
    },
    {
      "epoch": 6.88055704338511,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 28903
    },
    {
      "epoch": 6.880795096113789,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 28904
    },
    {
      "epoch": 6.8810331488424685,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8338,
      "step": 28905
    },
    {
      "epoch": 6.881271201571148,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 28906
    },
    {
      "epoch": 6.881509254299828,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 28907
    },
    {
      "epoch": 6.8817473070285065,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 28908
    },
    {
      "epoch": 6.881985359757186,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1821,
      "step": 28909
    },
    {
      "epoch": 6.882223412485866,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 28910
    },
    {
      "epoch": 6.882461465214545,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 28911
    },
    {
      "epoch": 6.882699517943225,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 28912
    },
    {
      "epoch": 6.882937570671904,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 28913
    },
    {
      "epoch": 6.883175623400583,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 28914
    },
    {
      "epoch": 6.883413676129263,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 28915
    },
    {
      "epoch": 6.883651728857942,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 28916
    },
    {
      "epoch": 6.883889781586621,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 28917
    },
    {
      "epoch": 6.884127834315301,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 28918
    },
    {
      "epoch": 6.88436588704398,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 28919
    },
    {
      "epoch": 6.88460393977266,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8823,
      "step": 28920
    },
    {
      "epoch": 6.884841992501339,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 28921
    },
    {
      "epoch": 6.885080045230018,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 28922
    },
    {
      "epoch": 6.885318097958698,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 28923
    },
    {
      "epoch": 6.8855561506873775,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 28924
    },
    {
      "epoch": 6.885794203416056,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 28925
    },
    {
      "epoch": 6.886032256144736,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 28926
    },
    {
      "epoch": 6.886270308873415,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 28927
    },
    {
      "epoch": 6.886508361602095,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 28928
    },
    {
      "epoch": 6.886746414330775,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8857,
      "step": 28929
    },
    {
      "epoch": 6.886984467059453,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 28930
    },
    {
      "epoch": 6.887222519788133,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 28931
    },
    {
      "epoch": 6.8874605725168125,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 28932
    },
    {
      "epoch": 6.887698625245492,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 28933
    },
    {
      "epoch": 6.887936677974171,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 28934
    },
    {
      "epoch": 6.8881747307028505,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2456,
      "step": 28935
    },
    {
      "epoch": 6.88841278343153,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3308,
      "step": 28936
    },
    {
      "epoch": 6.88865083616021,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 28937
    },
    {
      "epoch": 6.888888888888889,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 28938
    },
    {
      "epoch": 6.889126941617568,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.4153,
      "step": 28939
    },
    {
      "epoch": 6.889364994346248,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 28940
    },
    {
      "epoch": 6.889603047074927,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8288,
      "step": 28941
    },
    {
      "epoch": 6.889841099803607,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 28942
    },
    {
      "epoch": 6.8900791525322855,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 28943
    },
    {
      "epoch": 6.890317205260965,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3236,
      "step": 28944
    },
    {
      "epoch": 6.890555257989645,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 28945
    },
    {
      "epoch": 6.890793310718324,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 28946
    },
    {
      "epoch": 6.891031363447004,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 28947
    },
    {
      "epoch": 6.891269416175683,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 28948
    },
    {
      "epoch": 6.891507468904362,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 28949
    },
    {
      "epoch": 6.891745521633042,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 28950
    },
    {
      "epoch": 6.8919835743617215,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3972,
      "step": 28951
    },
    {
      "epoch": 6.8922216270904,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 28952
    },
    {
      "epoch": 6.89245967981908,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 28953
    },
    {
      "epoch": 6.892697732547759,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 28954
    },
    {
      "epoch": 6.892935785276439,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 28955
    },
    {
      "epoch": 6.893173838005119,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 28956
    },
    {
      "epoch": 6.893411890733797,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8194,
      "step": 28957
    },
    {
      "epoch": 6.893649943462477,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 28958
    },
    {
      "epoch": 6.8938879961911566,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 28959
    },
    {
      "epoch": 6.894126048919835,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 28960
    },
    {
      "epoch": 6.894364101648515,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 28961
    },
    {
      "epoch": 6.8946021543771945,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 28962
    },
    {
      "epoch": 6.894840207105874,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 28963
    },
    {
      "epoch": 6.895078259834554,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7584,
      "step": 28964
    },
    {
      "epoch": 6.895316312563232,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 28965
    },
    {
      "epoch": 6.895554365291912,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 28966
    },
    {
      "epoch": 6.895792418020592,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 28967
    },
    {
      "epoch": 6.896030470749271,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 28968
    },
    {
      "epoch": 6.89626852347795,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 28969
    },
    {
      "epoch": 6.8965065762066295,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 28970
    },
    {
      "epoch": 6.896744628935309,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 28971
    },
    {
      "epoch": 6.896982681663989,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3617,
      "step": 28972
    },
    {
      "epoch": 6.897220734392668,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 28973
    },
    {
      "epoch": 6.897458787121347,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 28974
    },
    {
      "epoch": 6.897696839850027,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2367,
      "step": 28975
    },
    {
      "epoch": 6.897934892578706,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 28976
    },
    {
      "epoch": 6.898172945307386,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 28977
    },
    {
      "epoch": 6.898410998036065,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 28978
    },
    {
      "epoch": 6.898649050764744,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 28979
    },
    {
      "epoch": 6.898887103493424,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 28980
    },
    {
      "epoch": 6.899125156222103,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0781,
      "step": 28981
    },
    {
      "epoch": 6.899363208950783,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3023,
      "step": 28982
    },
    {
      "epoch": 6.899601261679462,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7929,
      "step": 28983
    },
    {
      "epoch": 6.899839314408141,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 28984
    },
    {
      "epoch": 6.900077367136821,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 28985
    },
    {
      "epoch": 6.900315419865501,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2997,
      "step": 28986
    },
    {
      "epoch": 6.900553472594179,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 28987
    },
    {
      "epoch": 6.900791525322859,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 28988
    },
    {
      "epoch": 6.9010295780515385,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 28989
    },
    {
      "epoch": 6.901267630780218,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 28990
    },
    {
      "epoch": 6.901505683508898,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 28991
    },
    {
      "epoch": 6.901743736237576,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 28992
    },
    {
      "epoch": 6.901981788966256,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 28993
    },
    {
      "epoch": 6.902219841694936,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 28994
    },
    {
      "epoch": 6.902457894423614,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 28995
    },
    {
      "epoch": 6.902695947152294,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 28996
    },
    {
      "epoch": 6.902933999880974,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 28997
    },
    {
      "epoch": 6.903172052609653,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 28998
    },
    {
      "epoch": 6.903410105338333,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4074,
      "step": 28999
    },
    {
      "epoch": 6.9036481580670115,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 29000
    },
    {
      "epoch": 6.9036481580670115,
      "eval_runtime": 0.5225,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 29000
    },
    {
      "epoch": 6.903886210795691,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 29001
    },
    {
      "epoch": 6.904124263524371,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 29002
    },
    {
      "epoch": 6.90436231625305,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 29003
    },
    {
      "epoch": 6.904600368981729,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 29004
    },
    {
      "epoch": 6.904838421710409,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 29005
    },
    {
      "epoch": 6.905076474439088,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 29006
    },
    {
      "epoch": 6.905314527167768,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 29007
    },
    {
      "epoch": 6.905552579896447,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 29008
    },
    {
      "epoch": 6.905790632625126,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3083,
      "step": 29009
    },
    {
      "epoch": 6.906028685353806,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 29010
    },
    {
      "epoch": 6.906266738082485,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 29011
    },
    {
      "epoch": 6.906504790811165,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8331,
      "step": 29012
    },
    {
      "epoch": 6.906742843539844,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 29013
    },
    {
      "epoch": 6.906980896268523,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 29014
    },
    {
      "epoch": 6.907218948997203,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 29015
    },
    {
      "epoch": 6.9074570017258825,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 29016
    },
    {
      "epoch": 6.907695054454562,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 29017
    },
    {
      "epoch": 6.907933107183241,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 29018
    },
    {
      "epoch": 6.90817115991192,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 29019
    },
    {
      "epoch": 6.9084092126406,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 29020
    },
    {
      "epoch": 6.90864726536928,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 29021
    },
    {
      "epoch": 6.908885318097958,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7066,
      "step": 29022
    },
    {
      "epoch": 6.909123370826638,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 29023
    },
    {
      "epoch": 6.909361423555318,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 29024
    },
    {
      "epoch": 6.909599476283997,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 29025
    },
    {
      "epoch": 6.909837529012677,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 29026
    },
    {
      "epoch": 6.9100755817413555,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 29027
    },
    {
      "epoch": 6.910313634470035,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 29028
    },
    {
      "epoch": 6.910551687198715,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 29029
    },
    {
      "epoch": 6.910789739927393,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 29030
    },
    {
      "epoch": 6.911027792656073,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 29031
    },
    {
      "epoch": 6.911265845384753,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 29032
    },
    {
      "epoch": 6.911503898113432,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 29033
    },
    {
      "epoch": 6.911741950842112,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 29034
    },
    {
      "epoch": 6.911980003570791,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 29035
    },
    {
      "epoch": 6.91221805629947,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1977,
      "step": 29036
    },
    {
      "epoch": 6.91245610902815,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 29037
    },
    {
      "epoch": 6.912694161756829,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 29038
    },
    {
      "epoch": 6.912932214485508,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 29039
    },
    {
      "epoch": 6.913170267214188,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3039,
      "step": 29040
    },
    {
      "epoch": 6.913408319942867,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2835,
      "step": 29041
    },
    {
      "epoch": 6.913646372671547,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7533,
      "step": 29042
    },
    {
      "epoch": 6.9138844254002265,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 29043
    },
    {
      "epoch": 6.914122478128905,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7765,
      "step": 29044
    },
    {
      "epoch": 6.914360530857585,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 29045
    },
    {
      "epoch": 6.9145985835862644,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 29046
    },
    {
      "epoch": 6.914836636314944,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 29047
    },
    {
      "epoch": 6.915074689043623,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 29048
    },
    {
      "epoch": 6.915312741772302,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 29049
    },
    {
      "epoch": 6.915550794500982,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8561,
      "step": 29050
    },
    {
      "epoch": 6.915788847229662,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 29051
    },
    {
      "epoch": 6.916026899958341,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 29052
    },
    {
      "epoch": 6.91626495268702,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 29053
    },
    {
      "epoch": 6.9165030054156995,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 29054
    },
    {
      "epoch": 6.916741058144379,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 29055
    },
    {
      "epoch": 6.916979110873059,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 29056
    },
    {
      "epoch": 6.917217163601737,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9722,
      "step": 29057
    },
    {
      "epoch": 6.917455216330417,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 29058
    },
    {
      "epoch": 6.917693269059097,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 29059
    },
    {
      "epoch": 6.917931321787776,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 29060
    },
    {
      "epoch": 6.918169374516456,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6374,
      "step": 29061
    },
    {
      "epoch": 6.918407427245135,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 29062
    },
    {
      "epoch": 6.918645479973814,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 29063
    },
    {
      "epoch": 6.918883532702494,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 29064
    },
    {
      "epoch": 6.9191215854311725,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 29065
    },
    {
      "epoch": 6.919359638159852,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 29066
    },
    {
      "epoch": 6.919597690888532,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 29067
    },
    {
      "epoch": 6.919835743617211,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 29068
    },
    {
      "epoch": 6.920073796345891,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9895,
      "step": 29069
    },
    {
      "epoch": 6.92031184907457,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 29070
    },
    {
      "epoch": 6.920549901803249,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3715,
      "step": 29071
    },
    {
      "epoch": 6.920787954531929,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 29072
    },
    {
      "epoch": 6.9210260072606085,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 29073
    },
    {
      "epoch": 6.921264059989287,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 29074
    },
    {
      "epoch": 6.921502112717967,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 29075
    },
    {
      "epoch": 6.921740165446646,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 29076
    },
    {
      "epoch": 6.921978218175326,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 29077
    },
    {
      "epoch": 6.922216270904006,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 29078
    },
    {
      "epoch": 6.922454323632684,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 29079
    },
    {
      "epoch": 6.922692376361364,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 29080
    },
    {
      "epoch": 6.9229304290900435,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 29081
    },
    {
      "epoch": 6.923168481818723,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 29082
    },
    {
      "epoch": 6.923406534547402,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4999,
      "step": 29083
    },
    {
      "epoch": 6.9236445872760815,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2954,
      "step": 29084
    },
    {
      "epoch": 6.923882640004761,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 29085
    },
    {
      "epoch": 6.924120692733441,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9363,
      "step": 29086
    },
    {
      "epoch": 6.92435874546212,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 29087
    },
    {
      "epoch": 6.924596798190799,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 29088
    },
    {
      "epoch": 6.924834850919479,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 29089
    },
    {
      "epoch": 6.925072903648158,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 29090
    },
    {
      "epoch": 6.925310956376838,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 29091
    },
    {
      "epoch": 6.9255490091055165,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 29092
    },
    {
      "epoch": 6.925787061834196,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 29093
    },
    {
      "epoch": 6.926025114562876,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 29094
    },
    {
      "epoch": 6.926263167291555,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 29095
    },
    {
      "epoch": 6.926501220020235,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 29096
    },
    {
      "epoch": 6.926739272748914,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 29097
    },
    {
      "epoch": 6.926977325477593,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4651,
      "step": 29098
    },
    {
      "epoch": 6.927215378206273,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 29099
    },
    {
      "epoch": 6.9274534309349525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 29100
    },
    {
      "epoch": 6.9274534309349525,
      "eval_runtime": 0.5348,
      "eval_samples_per_second": 1.87,
      "eval_steps_per_second": 1.87,
      "step": 29100
    },
    {
      "epoch": 6.927691483663631,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 29101
    },
    {
      "epoch": 6.927929536392311,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 29102
    },
    {
      "epoch": 6.92816758912099,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 29103
    },
    {
      "epoch": 6.92840564184967,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 29104
    },
    {
      "epoch": 6.928643694578349,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 29105
    },
    {
      "epoch": 6.928881747307028,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4086,
      "step": 29106
    },
    {
      "epoch": 6.929119800035708,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 29107
    },
    {
      "epoch": 6.9293578527643875,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 29108
    },
    {
      "epoch": 6.929595905493066,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 29109
    },
    {
      "epoch": 6.929833958221746,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 29110
    },
    {
      "epoch": 6.9300720109504255,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 29111
    },
    {
      "epoch": 6.930310063679105,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 29112
    },
    {
      "epoch": 6.930548116407785,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7556,
      "step": 29113
    },
    {
      "epoch": 6.930786169136463,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 29114
    },
    {
      "epoch": 6.931024221865143,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 29115
    },
    {
      "epoch": 6.931262274593823,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 29116
    },
    {
      "epoch": 6.931500327322502,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 29117
    },
    {
      "epoch": 6.931738380051181,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 29118
    },
    {
      "epoch": 6.9319764327798605,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 29119
    },
    {
      "epoch": 6.93221448550854,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 29120
    },
    {
      "epoch": 6.93245253823722,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 29121
    },
    {
      "epoch": 6.932690590965899,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9533,
      "step": 29122
    },
    {
      "epoch": 6.932928643694578,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 29123
    },
    {
      "epoch": 6.933166696423258,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.176,
      "step": 29124
    },
    {
      "epoch": 6.933404749151937,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 29125
    },
    {
      "epoch": 6.933642801880617,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 29126
    },
    {
      "epoch": 6.933880854609296,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 29127
    },
    {
      "epoch": 6.934118907337975,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2003,
      "step": 29128
    },
    {
      "epoch": 6.934356960066655,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 29129
    },
    {
      "epoch": 6.934595012795334,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 29130
    },
    {
      "epoch": 6.934833065524014,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 29131
    },
    {
      "epoch": 6.935071118252693,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 29132
    },
    {
      "epoch": 6.935309170981372,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 29133
    },
    {
      "epoch": 6.935547223710052,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 29134
    },
    {
      "epoch": 6.9357852764387316,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 29135
    },
    {
      "epoch": 6.93602332916741,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 29136
    },
    {
      "epoch": 6.93626138189609,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7912,
      "step": 29137
    },
    {
      "epoch": 6.9364994346247695,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 29138
    },
    {
      "epoch": 6.936737487353449,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 29139
    },
    {
      "epoch": 6.936975540082129,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 29140
    },
    {
      "epoch": 6.937213592810807,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.3494,
      "step": 29141
    },
    {
      "epoch": 6.937451645539487,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 29142
    },
    {
      "epoch": 6.937689698268167,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 29143
    },
    {
      "epoch": 6.937927750996845,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 29144
    },
    {
      "epoch": 6.938165803725525,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 29145
    },
    {
      "epoch": 6.9384038564542045,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 29146
    },
    {
      "epoch": 6.938641909182884,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1233,
      "step": 29147
    },
    {
      "epoch": 6.938879961911564,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 29148
    },
    {
      "epoch": 6.9391180146402425,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 29149
    },
    {
      "epoch": 6.939356067368922,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2566,
      "step": 29150
    },
    {
      "epoch": 6.939594120097602,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 29151
    },
    {
      "epoch": 6.939832172826281,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 29152
    },
    {
      "epoch": 6.94007022555496,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 29153
    },
    {
      "epoch": 6.94030827828364,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 29154
    },
    {
      "epoch": 6.940546331012319,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 29155
    },
    {
      "epoch": 6.940784383740999,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 29156
    },
    {
      "epoch": 6.941022436469678,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 29157
    },
    {
      "epoch": 6.941260489198357,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 29158
    },
    {
      "epoch": 6.941498541927037,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 29159
    },
    {
      "epoch": 6.941736594655716,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 29160
    },
    {
      "epoch": 6.941974647384396,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 29161
    },
    {
      "epoch": 6.942212700113075,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3714,
      "step": 29162
    },
    {
      "epoch": 6.942450752841754,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 29163
    },
    {
      "epoch": 6.942688805570434,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 29164
    },
    {
      "epoch": 6.9429268582991135,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 29165
    },
    {
      "epoch": 6.943164911027793,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8044,
      "step": 29166
    },
    {
      "epoch": 6.943402963756472,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 29167
    },
    {
      "epoch": 6.943641016485151,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.4148,
      "step": 29168
    },
    {
      "epoch": 6.943879069213831,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 29169
    },
    {
      "epoch": 6.944117121942511,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7721,
      "step": 29170
    },
    {
      "epoch": 6.944355174671189,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 29171
    },
    {
      "epoch": 6.944593227399869,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 29172
    },
    {
      "epoch": 6.944831280128549,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 29173
    },
    {
      "epoch": 6.945069332857228,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 29174
    },
    {
      "epoch": 6.945307385585908,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 29175
    },
    {
      "epoch": 6.9455454383145865,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 29176
    },
    {
      "epoch": 6.945783491043266,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 29177
    },
    {
      "epoch": 6.946021543771946,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 29178
    },
    {
      "epoch": 6.946259596500624,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 29179
    },
    {
      "epoch": 6.946497649229304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 29180
    },
    {
      "epoch": 6.946735701957984,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 29181
    },
    {
      "epoch": 6.946973754686663,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 29182
    },
    {
      "epoch": 6.947211807415343,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 29183
    },
    {
      "epoch": 6.9474498601440215,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3675,
      "step": 29184
    },
    {
      "epoch": 6.947687912872701,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 29185
    },
    {
      "epoch": 6.947925965601381,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 29186
    },
    {
      "epoch": 6.94816401833006,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9933,
      "step": 29187
    },
    {
      "epoch": 6.948402071058739,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.051,
      "step": 29188
    },
    {
      "epoch": 6.948640123787419,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 29189
    },
    {
      "epoch": 6.948878176516098,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 29190
    },
    {
      "epoch": 6.949116229244778,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1466,
      "step": 29191
    },
    {
      "epoch": 6.9493542819734575,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 29192
    },
    {
      "epoch": 6.949592334702136,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 29193
    },
    {
      "epoch": 6.949830387430816,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 29194
    },
    {
      "epoch": 6.950068440159495,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 29195
    },
    {
      "epoch": 6.950306492888175,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 29196
    },
    {
      "epoch": 6.950544545616854,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 29197
    },
    {
      "epoch": 6.950782598345533,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0824,
      "step": 29198
    },
    {
      "epoch": 6.951020651074213,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3299,
      "step": 29199
    },
    {
      "epoch": 6.951258703802893,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 29200
    },
    {
      "epoch": 6.951258703802893,
      "eval_runtime": 0.5224,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 29200
    },
    {
      "epoch": 6.951496756531572,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1178,
      "step": 29201
    },
    {
      "epoch": 6.951734809260251,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 29202
    },
    {
      "epoch": 6.9519728619889305,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 29203
    },
    {
      "epoch": 6.95221091471761,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 29204
    },
    {
      "epoch": 6.95244896744629,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 29205
    },
    {
      "epoch": 6.952687020174968,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 29206
    },
    {
      "epoch": 6.952925072903648,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3607,
      "step": 29207
    },
    {
      "epoch": 6.953163125632328,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 29208
    },
    {
      "epoch": 6.953401178361007,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 29209
    },
    {
      "epoch": 6.953639231089687,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 29210
    },
    {
      "epoch": 6.953877283818366,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 29211
    },
    {
      "epoch": 6.954115336547045,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 29212
    },
    {
      "epoch": 6.954353389275725,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 29213
    },
    {
      "epoch": 6.9545914420044035,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0372,
      "step": 29214
    },
    {
      "epoch": 6.954829494733083,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6871,
      "step": 29215
    },
    {
      "epoch": 6.955067547461763,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 29216
    },
    {
      "epoch": 6.955305600190442,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 29217
    },
    {
      "epoch": 6.955543652919122,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 29218
    },
    {
      "epoch": 6.955781705647801,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 29219
    },
    {
      "epoch": 6.95601975837648,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 29220
    },
    {
      "epoch": 6.95625781110516,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3289,
      "step": 29221
    },
    {
      "epoch": 6.956495863833839,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 29222
    },
    {
      "epoch": 6.956733916562518,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8621,
      "step": 29223
    },
    {
      "epoch": 6.956971969291198,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 29224
    },
    {
      "epoch": 6.957210022019877,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 29225
    },
    {
      "epoch": 6.957448074748557,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 29226
    },
    {
      "epoch": 6.957686127477237,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9442,
      "step": 29227
    },
    {
      "epoch": 6.957924180205915,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2309,
      "step": 29228
    },
    {
      "epoch": 6.958162232934595,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 29229
    },
    {
      "epoch": 6.9584002856632745,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 29230
    },
    {
      "epoch": 6.958638338391954,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 29231
    },
    {
      "epoch": 6.958876391120633,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 29232
    },
    {
      "epoch": 6.959114443849312,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3068,
      "step": 29233
    },
    {
      "epoch": 6.959352496577992,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7425,
      "step": 29234
    },
    {
      "epoch": 6.959590549306672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 29235
    },
    {
      "epoch": 6.959828602035351,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 29236
    },
    {
      "epoch": 6.96006665476403,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.4639,
      "step": 29237
    },
    {
      "epoch": 6.96030470749271,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 29238
    },
    {
      "epoch": 6.960542760221389,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3871,
      "step": 29239
    },
    {
      "epoch": 6.960780812950069,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 29240
    },
    {
      "epoch": 6.9610188656787475,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 29241
    },
    {
      "epoch": 6.961256918407427,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 29242
    },
    {
      "epoch": 6.961494971136107,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 29243
    },
    {
      "epoch": 6.961733023864786,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 29244
    },
    {
      "epoch": 6.961971076593466,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3571,
      "step": 29245
    },
    {
      "epoch": 6.962209129322145,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 29246
    },
    {
      "epoch": 6.962447182050824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 29247
    },
    {
      "epoch": 6.962685234779504,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9113,
      "step": 29248
    },
    {
      "epoch": 6.962923287508183,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3149,
      "step": 29249
    },
    {
      "epoch": 6.963161340236862,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 29250
    },
    {
      "epoch": 6.963399392965542,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 29251
    },
    {
      "epoch": 6.963637445694221,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 29252
    },
    {
      "epoch": 6.963875498422901,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 29253
    },
    {
      "epoch": 6.96411355115158,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 29254
    },
    {
      "epoch": 6.964351603880259,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0838,
      "step": 29255
    },
    {
      "epoch": 6.964589656608939,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 29256
    },
    {
      "epoch": 6.9648277093376185,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 29257
    },
    {
      "epoch": 6.965065762066297,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 29258
    },
    {
      "epoch": 6.965303814794977,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 29259
    },
    {
      "epoch": 6.9655418675236564,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 29260
    },
    {
      "epoch": 6.965779920252336,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 29261
    },
    {
      "epoch": 6.966017972981016,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2234,
      "step": 29262
    },
    {
      "epoch": 6.966256025709694,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 29263
    },
    {
      "epoch": 6.966494078438374,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 29264
    },
    {
      "epoch": 6.966732131167054,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 29265
    },
    {
      "epoch": 6.966970183895733,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 29266
    },
    {
      "epoch": 6.967208236624412,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7815,
      "step": 29267
    },
    {
      "epoch": 6.9674462893530915,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1389,
      "step": 29268
    },
    {
      "epoch": 6.967684342081771,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 29269
    },
    {
      "epoch": 6.967922394810451,
      "grad_norm": 6.5625,
      "learning_rate": 3e-05,
      "loss": 2.3887,
      "step": 29270
    },
    {
      "epoch": 6.96816044753913,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 29271
    },
    {
      "epoch": 6.968398500267809,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7967,
      "step": 29272
    },
    {
      "epoch": 6.968636552996489,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 29273
    },
    {
      "epoch": 6.968874605725168,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 29274
    },
    {
      "epoch": 6.969112658453848,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 29275
    },
    {
      "epoch": 6.969350711182527,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 29276
    },
    {
      "epoch": 6.969588763911206,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 29277
    },
    {
      "epoch": 6.969826816639886,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 29278
    },
    {
      "epoch": 6.970064869368565,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 29279
    },
    {
      "epoch": 6.970302922097245,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7263,
      "step": 29280
    },
    {
      "epoch": 6.970540974825924,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9942,
      "step": 29281
    },
    {
      "epoch": 6.970779027554603,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 29282
    },
    {
      "epoch": 6.971017080283283,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 29283
    },
    {
      "epoch": 6.9712551330119625,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 29284
    },
    {
      "epoch": 6.971493185740641,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 29285
    },
    {
      "epoch": 6.971731238469321,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 29286
    },
    {
      "epoch": 6.9719692911980005,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 29287
    },
    {
      "epoch": 6.97220734392668,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 29288
    },
    {
      "epoch": 6.972445396655359,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 29289
    },
    {
      "epoch": 6.972683449384038,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 29290
    },
    {
      "epoch": 6.972921502112718,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 29291
    },
    {
      "epoch": 6.973159554841398,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 29292
    },
    {
      "epoch": 6.973397607570076,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1383,
      "step": 29293
    },
    {
      "epoch": 6.973635660298756,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 29294
    },
    {
      "epoch": 6.9738737130274355,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 29295
    },
    {
      "epoch": 6.974111765756115,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 29296
    },
    {
      "epoch": 6.974349818484795,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 29297
    },
    {
      "epoch": 6.9745878712134735,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8623,
      "step": 29298
    },
    {
      "epoch": 6.974825923942153,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 29299
    },
    {
      "epoch": 6.975063976670833,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 29300
    },
    {
      "epoch": 6.975063976670833,
      "eval_runtime": 0.5268,
      "eval_samples_per_second": 1.898,
      "eval_steps_per_second": 1.898,
      "step": 29300
    },
    {
      "epoch": 6.975302029399512,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 29301
    },
    {
      "epoch": 6.975540082128191,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.165,
      "step": 29302
    },
    {
      "epoch": 6.975778134856871,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 29303
    },
    {
      "epoch": 6.97601618758555,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 29304
    },
    {
      "epoch": 6.97625424031423,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 29305
    },
    {
      "epoch": 6.976492293042909,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3322,
      "step": 29306
    },
    {
      "epoch": 6.976730345771588,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 29307
    },
    {
      "epoch": 6.976968398500268,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 29308
    },
    {
      "epoch": 6.977206451228947,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0459,
      "step": 29309
    },
    {
      "epoch": 6.977444503957627,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 29310
    },
    {
      "epoch": 6.977682556686306,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 29311
    },
    {
      "epoch": 6.977920609414985,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 29312
    },
    {
      "epoch": 6.978158662143665,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 29313
    },
    {
      "epoch": 6.9783967148723445,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4584,
      "step": 29314
    },
    {
      "epoch": 6.978634767601024,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 29315
    },
    {
      "epoch": 6.978872820329703,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 29316
    },
    {
      "epoch": 6.979110873058382,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2559,
      "step": 29317
    },
    {
      "epoch": 6.979348925787062,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4397,
      "step": 29318
    },
    {
      "epoch": 6.979586978515742,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 29319
    },
    {
      "epoch": 6.97982503124442,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 29320
    },
    {
      "epoch": 6.9800630839731,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.293,
      "step": 29321
    },
    {
      "epoch": 6.9803011367017795,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 29322
    },
    {
      "epoch": 6.980539189430459,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3925,
      "step": 29323
    },
    {
      "epoch": 6.980777242159138,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.4303,
      "step": 29324
    },
    {
      "epoch": 6.9810152948878175,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 29325
    },
    {
      "epoch": 6.981253347616497,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 29326
    },
    {
      "epoch": 6.981491400345177,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 29327
    },
    {
      "epoch": 6.981729453073855,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 29328
    },
    {
      "epoch": 6.981967505802535,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 29329
    },
    {
      "epoch": 6.982205558531215,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 29330
    },
    {
      "epoch": 6.982443611259894,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 29331
    },
    {
      "epoch": 6.982681663988574,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 29332
    },
    {
      "epoch": 6.9829197167172525,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1485,
      "step": 29333
    },
    {
      "epoch": 6.983157769445932,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 29334
    },
    {
      "epoch": 6.983395822174612,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 29335
    },
    {
      "epoch": 6.983633874903291,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 29336
    },
    {
      "epoch": 6.98387192763197,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 29337
    },
    {
      "epoch": 6.98410998036065,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 29338
    },
    {
      "epoch": 6.984348033089329,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 29339
    },
    {
      "epoch": 6.984586085818009,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 29340
    },
    {
      "epoch": 6.9848241385466885,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3773,
      "step": 29341
    },
    {
      "epoch": 6.985062191275367,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8339,
      "step": 29342
    },
    {
      "epoch": 6.985300244004047,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 29343
    },
    {
      "epoch": 6.985538296732726,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 29344
    },
    {
      "epoch": 6.985776349461406,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 29345
    },
    {
      "epoch": 6.986014402190085,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 29346
    },
    {
      "epoch": 6.986252454918764,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 29347
    },
    {
      "epoch": 6.986490507647444,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 29348
    },
    {
      "epoch": 6.9867285603761236,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 29349
    },
    {
      "epoch": 6.986966613104803,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 29350
    },
    {
      "epoch": 6.987204665833482,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 29351
    },
    {
      "epoch": 6.9874427185621615,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 29352
    },
    {
      "epoch": 6.987680771290841,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 29353
    },
    {
      "epoch": 6.987918824019521,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 29354
    },
    {
      "epoch": 6.988156876748199,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 29355
    },
    {
      "epoch": 6.988394929476879,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 29356
    },
    {
      "epoch": 6.988632982205559,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 29357
    },
    {
      "epoch": 6.988871034934238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 29358
    },
    {
      "epoch": 6.989109087662918,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 29359
    },
    {
      "epoch": 6.9893471403915965,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 29360
    },
    {
      "epoch": 6.989585193120276,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 29361
    },
    {
      "epoch": 6.989823245848956,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 29362
    },
    {
      "epoch": 6.9900612985776345,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7326,
      "step": 29363
    },
    {
      "epoch": 6.990299351306314,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 29364
    },
    {
      "epoch": 6.990537404034994,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 29365
    },
    {
      "epoch": 6.990775456763673,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 29366
    },
    {
      "epoch": 6.991013509492353,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 29367
    },
    {
      "epoch": 6.991251562221032,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 29368
    },
    {
      "epoch": 6.991489614949711,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 29369
    },
    {
      "epoch": 6.991727667678391,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 29370
    },
    {
      "epoch": 6.99196572040707,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7577,
      "step": 29371
    },
    {
      "epoch": 6.992203773135749,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 29372
    },
    {
      "epoch": 6.992441825864429,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3292,
      "step": 29373
    },
    {
      "epoch": 6.992679878593108,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 29374
    },
    {
      "epoch": 6.992917931321788,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 29375
    },
    {
      "epoch": 6.993155984050468,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1293,
      "step": 29376
    },
    {
      "epoch": 6.993394036779146,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 29377
    },
    {
      "epoch": 6.993632089507826,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 29378
    },
    {
      "epoch": 6.9938701422365055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8391,
      "step": 29379
    },
    {
      "epoch": 6.994108194965185,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 29380
    },
    {
      "epoch": 6.994346247693864,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 29381
    },
    {
      "epoch": 6.994584300422543,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 29382
    },
    {
      "epoch": 6.994822353151223,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 29383
    },
    {
      "epoch": 6.995060405879903,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.4192,
      "step": 29384
    },
    {
      "epoch": 6.995298458608582,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 29385
    },
    {
      "epoch": 6.995536511337261,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7217,
      "step": 29386
    },
    {
      "epoch": 6.995774564065941,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 29387
    },
    {
      "epoch": 6.99601261679462,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 29388
    },
    {
      "epoch": 6.9962506695233,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2532,
      "step": 29389
    },
    {
      "epoch": 6.9964887222519785,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 29390
    },
    {
      "epoch": 6.996726774980658,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 29391
    },
    {
      "epoch": 6.996964827709338,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 29392
    },
    {
      "epoch": 6.997202880438017,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 29393
    },
    {
      "epoch": 6.997440933166697,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 29394
    },
    {
      "epoch": 6.997678985895376,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 29395
    },
    {
      "epoch": 6.997917038624055,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 29396
    },
    {
      "epoch": 6.998155091352735,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 29397
    },
    {
      "epoch": 6.9983931440814136,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9977,
      "step": 29398
    },
    {
      "epoch": 6.998631196810093,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 29399
    },
    {
      "epoch": 6.998869249538773,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 29400
    },
    {
      "epoch": 6.998869249538773,
      "eval_runtime": 0.5367,
      "eval_samples_per_second": 1.863,
      "eval_steps_per_second": 1.863,
      "step": 29400
    },
    {
      "epoch": 6.999107302267452,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 29401
    },
    {
      "epoch": 6.999345354996132,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 29402
    },
    {
      "epoch": 6.999583407724811,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7185,
      "step": 29403
    },
    {
      "epoch": 6.99982146045349,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 29404
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 29405
    },
    {
      "epoch": 7.00023805272868,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 29406
    },
    {
      "epoch": 7.000476105457359,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 29407
    },
    {
      "epoch": 7.000714158186038,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3901,
      "step": 29408
    },
    {
      "epoch": 7.0009522109147175,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 29409
    },
    {
      "epoch": 7.001190263643397,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 29410
    },
    {
      "epoch": 7.001428316372077,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 29411
    },
    {
      "epoch": 7.0016663691007555,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 29412
    },
    {
      "epoch": 7.001904421829435,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 29413
    },
    {
      "epoch": 7.002142474558115,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 29414
    },
    {
      "epoch": 7.002380527286794,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 29415
    },
    {
      "epoch": 7.002618580015473,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 29416
    },
    {
      "epoch": 7.002856632744153,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7739,
      "step": 29417
    },
    {
      "epoch": 7.003094685472832,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 29418
    },
    {
      "epoch": 7.003332738201512,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 29419
    },
    {
      "epoch": 7.003570790930191,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 29420
    },
    {
      "epoch": 7.00380884365887,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 29421
    },
    {
      "epoch": 7.00404689638755,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 29422
    },
    {
      "epoch": 7.004284949116229,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 29423
    },
    {
      "epoch": 7.004523001844909,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 29424
    },
    {
      "epoch": 7.004761054573588,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 29425
    },
    {
      "epoch": 7.004999107302267,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 29426
    },
    {
      "epoch": 7.005237160030947,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 29427
    },
    {
      "epoch": 7.0054752127596265,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 29428
    },
    {
      "epoch": 7.005713265488306,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 29429
    },
    {
      "epoch": 7.005951318216985,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 29430
    },
    {
      "epoch": 7.006189370945664,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2883,
      "step": 29431
    },
    {
      "epoch": 7.006427423674344,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 29432
    },
    {
      "epoch": 7.006665476403024,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 29433
    },
    {
      "epoch": 7.006903529131702,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 29434
    },
    {
      "epoch": 7.007141581860382,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 29435
    },
    {
      "epoch": 7.0073796345890615,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 29436
    },
    {
      "epoch": 7.007617687317741,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 29437
    },
    {
      "epoch": 7.00785574004642,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 29438
    },
    {
      "epoch": 7.0080937927750995,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 29439
    },
    {
      "epoch": 7.008331845503779,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.229,
      "step": 29440
    },
    {
      "epoch": 7.008569898232459,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 29441
    },
    {
      "epoch": 7.008807950961138,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6854,
      "step": 29442
    },
    {
      "epoch": 7.009046003689817,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 29443
    },
    {
      "epoch": 7.009284056418497,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 29444
    },
    {
      "epoch": 7.009522109147176,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 29445
    },
    {
      "epoch": 7.009760161875856,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.178,
      "step": 29446
    },
    {
      "epoch": 7.0099982146045345,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7873,
      "step": 29447
    },
    {
      "epoch": 7.010236267333214,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 29448
    },
    {
      "epoch": 7.010474320061894,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 29449
    },
    {
      "epoch": 7.010712372790573,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 29450
    },
    {
      "epoch": 7.010950425519252,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7197,
      "step": 29451
    },
    {
      "epoch": 7.011188478247932,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 29452
    },
    {
      "epoch": 7.011426530976611,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 29453
    },
    {
      "epoch": 7.011664583705291,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 29454
    },
    {
      "epoch": 7.0119026364339705,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 29455
    },
    {
      "epoch": 7.012140689162649,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 29456
    },
    {
      "epoch": 7.012378741891329,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2657,
      "step": 29457
    },
    {
      "epoch": 7.012616794620008,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 29458
    },
    {
      "epoch": 7.012854847348688,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 29459
    },
    {
      "epoch": 7.013092900077367,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 29460
    },
    {
      "epoch": 7.013330952806046,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 29461
    },
    {
      "epoch": 7.013569005534726,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 29462
    },
    {
      "epoch": 7.013807058263406,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 29463
    },
    {
      "epoch": 7.014045110992085,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 29464
    },
    {
      "epoch": 7.014283163720764,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 29465
    },
    {
      "epoch": 7.0145212164494435,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2119,
      "step": 29466
    },
    {
      "epoch": 7.014759269178123,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2132,
      "step": 29467
    },
    {
      "epoch": 7.014997321906803,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 29468
    },
    {
      "epoch": 7.015235374635481,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 29469
    },
    {
      "epoch": 7.015473427364161,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 29470
    },
    {
      "epoch": 7.015711480092841,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6998,
      "step": 29471
    },
    {
      "epoch": 7.01594953282152,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 29472
    },
    {
      "epoch": 7.016187585550199,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 29473
    },
    {
      "epoch": 7.0164256382788786,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 29474
    },
    {
      "epoch": 7.016663691007558,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 29475
    },
    {
      "epoch": 7.016901743736238,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 29476
    },
    {
      "epoch": 7.017139796464917,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 29477
    },
    {
      "epoch": 7.017377849193596,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 29478
    },
    {
      "epoch": 7.017615901922276,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 29479
    },
    {
      "epoch": 7.017853954650955,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.43,
      "step": 29480
    },
    {
      "epoch": 7.018092007379635,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 29481
    },
    {
      "epoch": 7.018330060108314,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 29482
    },
    {
      "epoch": 7.018568112836993,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 29483
    },
    {
      "epoch": 7.018806165565673,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3684,
      "step": 29484
    },
    {
      "epoch": 7.019044218294352,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 29485
    },
    {
      "epoch": 7.019282271023032,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 29486
    },
    {
      "epoch": 7.019520323751711,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 29487
    },
    {
      "epoch": 7.01975837648039,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 29488
    },
    {
      "epoch": 7.01999642920907,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 29489
    },
    {
      "epoch": 7.02023448193775,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1709,
      "step": 29490
    },
    {
      "epoch": 7.020472534666428,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 29491
    },
    {
      "epoch": 7.020710587395108,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 29492
    },
    {
      "epoch": 7.0209486401237875,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 29493
    },
    {
      "epoch": 7.021186692852467,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 29494
    },
    {
      "epoch": 7.021424745581146,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 29495
    },
    {
      "epoch": 7.021662798309825,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 29496
    },
    {
      "epoch": 7.021900851038505,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 29497
    },
    {
      "epoch": 7.022138903767185,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 29498
    },
    {
      "epoch": 7.022376956495864,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 29499
    },
    {
      "epoch": 7.022615009224543,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 29500
    },
    {
      "epoch": 7.022615009224543,
      "eval_runtime": 0.5182,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 29500
    },
    {
      "epoch": 7.022853061953223,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 29501
    },
    {
      "epoch": 7.023091114681902,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 29502
    },
    {
      "epoch": 7.023329167410582,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 29503
    },
    {
      "epoch": 7.0235672201392605,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 29504
    },
    {
      "epoch": 7.02380527286794,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 29505
    },
    {
      "epoch": 7.02404332559662,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 29506
    },
    {
      "epoch": 7.024281378325299,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 29507
    },
    {
      "epoch": 7.024519431053978,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 29508
    },
    {
      "epoch": 7.024757483782658,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 29509
    },
    {
      "epoch": 7.024995536511337,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 29510
    },
    {
      "epoch": 7.025233589240017,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 29511
    },
    {
      "epoch": 7.0254716419686964,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.122,
      "step": 29512
    },
    {
      "epoch": 7.025709694697375,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.696,
      "step": 29513
    },
    {
      "epoch": 7.025947747426055,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 29514
    },
    {
      "epoch": 7.026185800154734,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 29515
    },
    {
      "epoch": 7.026423852883414,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 29516
    },
    {
      "epoch": 7.026661905612093,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 29517
    },
    {
      "epoch": 7.026899958340772,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 29518
    },
    {
      "epoch": 7.027138011069452,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 29519
    },
    {
      "epoch": 7.0273760637981315,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 29520
    },
    {
      "epoch": 7.027614116526811,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 29521
    },
    {
      "epoch": 7.02785216925549,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 29522
    },
    {
      "epoch": 7.028090221984169,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9377,
      "step": 29523
    },
    {
      "epoch": 7.028328274712849,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 29524
    },
    {
      "epoch": 7.028566327441529,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 29525
    },
    {
      "epoch": 7.028804380170207,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 29526
    },
    {
      "epoch": 7.029042432898887,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 29527
    },
    {
      "epoch": 7.029280485627567,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 29528
    },
    {
      "epoch": 7.029518538356246,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 29529
    },
    {
      "epoch": 7.029756591084925,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2343,
      "step": 29530
    },
    {
      "epoch": 7.0299946438136045,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 29531
    },
    {
      "epoch": 7.030232696542284,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 29532
    },
    {
      "epoch": 7.030470749270964,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7819,
      "step": 29533
    },
    {
      "epoch": 7.030708801999643,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 29534
    },
    {
      "epoch": 7.030946854728322,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 29535
    },
    {
      "epoch": 7.031184907457002,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 29536
    },
    {
      "epoch": 7.031422960185681,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 29537
    },
    {
      "epoch": 7.031661012914361,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 29538
    },
    {
      "epoch": 7.03189906564304,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 29539
    },
    {
      "epoch": 7.032137118371719,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1002,
      "step": 29540
    },
    {
      "epoch": 7.032375171100399,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 29541
    },
    {
      "epoch": 7.032613223829078,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 29542
    },
    {
      "epoch": 7.032851276557757,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 29543
    },
    {
      "epoch": 7.033089329286437,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 29544
    },
    {
      "epoch": 7.033327382015116,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 29545
    },
    {
      "epoch": 7.033565434743796,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 29546
    },
    {
      "epoch": 7.0338034874724755,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 29547
    },
    {
      "epoch": 7.034041540201154,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 29548
    },
    {
      "epoch": 7.034279592929834,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3221,
      "step": 29549
    },
    {
      "epoch": 7.0345176456585135,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 29550
    },
    {
      "epoch": 7.034755698387193,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 29551
    },
    {
      "epoch": 7.034993751115872,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 29552
    },
    {
      "epoch": 7.035231803844551,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 29553
    },
    {
      "epoch": 7.035469856573231,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 29554
    },
    {
      "epoch": 7.035707909301911,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 29555
    },
    {
      "epoch": 7.03594596203059,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.198,
      "step": 29556
    },
    {
      "epoch": 7.036184014759269,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 29557
    },
    {
      "epoch": 7.0364220674879485,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 29558
    },
    {
      "epoch": 7.036660120216628,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 29559
    },
    {
      "epoch": 7.036898172945308,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 29560
    },
    {
      "epoch": 7.0371362256739864,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2346,
      "step": 29561
    },
    {
      "epoch": 7.037374278402666,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 29562
    },
    {
      "epoch": 7.037612331131346,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 29563
    },
    {
      "epoch": 7.037850383860025,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 29564
    },
    {
      "epoch": 7.038088436588704,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 29565
    },
    {
      "epoch": 7.038326489317384,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 29566
    },
    {
      "epoch": 7.038564542046063,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1235,
      "step": 29567
    },
    {
      "epoch": 7.038802594774743,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 2.3707,
      "step": 29568
    },
    {
      "epoch": 7.039040647503422,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 29569
    },
    {
      "epoch": 7.039278700232101,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1912,
      "step": 29570
    },
    {
      "epoch": 7.039516752960781,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 29571
    },
    {
      "epoch": 7.03975480568946,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3393,
      "step": 29572
    },
    {
      "epoch": 7.03999285841814,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 29573
    },
    {
      "epoch": 7.040230911146819,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 29574
    },
    {
      "epoch": 7.040468963875498,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 29575
    },
    {
      "epoch": 7.040707016604178,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 29576
    },
    {
      "epoch": 7.0409450693328575,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 29577
    },
    {
      "epoch": 7.041183122061537,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 29578
    },
    {
      "epoch": 7.041421174790216,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 29579
    },
    {
      "epoch": 7.041659227518895,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2515,
      "step": 29580
    },
    {
      "epoch": 7.041897280247575,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 29581
    },
    {
      "epoch": 7.042135332976255,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 29582
    },
    {
      "epoch": 7.042373385704933,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 29583
    },
    {
      "epoch": 7.042611438433613,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 29584
    },
    {
      "epoch": 7.0428494911622925,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 29585
    },
    {
      "epoch": 7.043087543890972,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2889,
      "step": 29586
    },
    {
      "epoch": 7.043325596619651,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 29587
    },
    {
      "epoch": 7.0435636493483305,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8988,
      "step": 29588
    },
    {
      "epoch": 7.04380170207701,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 29589
    },
    {
      "epoch": 7.04403975480569,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 29590
    },
    {
      "epoch": 7.044277807534369,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7012,
      "step": 29591
    },
    {
      "epoch": 7.044515860263048,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3728,
      "step": 29592
    },
    {
      "epoch": 7.044753912991728,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 29593
    },
    {
      "epoch": 7.044991965720407,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 29594
    },
    {
      "epoch": 7.045230018449087,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 29595
    },
    {
      "epoch": 7.0454680711777655,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1521,
      "step": 29596
    },
    {
      "epoch": 7.045706123906445,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 29597
    },
    {
      "epoch": 7.045944176635125,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 29598
    },
    {
      "epoch": 7.046182229363804,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0942,
      "step": 29599
    },
    {
      "epoch": 7.046420282092483,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 29600
    },
    {
      "epoch": 7.046420282092483,
      "eval_runtime": 0.5265,
      "eval_samples_per_second": 1.899,
      "eval_steps_per_second": 1.899,
      "step": 29600
    },
    {
      "epoch": 7.046658334821163,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 29601
    },
    {
      "epoch": 7.046896387549842,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 29602
    },
    {
      "epoch": 7.047134440278522,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 29603
    },
    {
      "epoch": 7.0473724930072015,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 29604
    },
    {
      "epoch": 7.04761054573588,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 29605
    },
    {
      "epoch": 7.04784859846456,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 29606
    },
    {
      "epoch": 7.048086651193239,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9885,
      "step": 29607
    },
    {
      "epoch": 7.048324703921919,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 29608
    },
    {
      "epoch": 7.048562756650598,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 29609
    },
    {
      "epoch": 7.048800809379277,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 29610
    },
    {
      "epoch": 7.049038862107957,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4813,
      "step": 29611
    },
    {
      "epoch": 7.0492769148366365,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 29612
    },
    {
      "epoch": 7.049514967565316,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 29613
    },
    {
      "epoch": 7.049753020293995,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 29614
    },
    {
      "epoch": 7.0499910730226745,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3165,
      "step": 29615
    },
    {
      "epoch": 7.050229125751354,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 29616
    },
    {
      "epoch": 7.050467178480034,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 29617
    },
    {
      "epoch": 7.050705231208712,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 29618
    },
    {
      "epoch": 7.050943283937392,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 29619
    },
    {
      "epoch": 7.051181336666072,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 29620
    },
    {
      "epoch": 7.051419389394751,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 29621
    },
    {
      "epoch": 7.05165744212343,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 29622
    },
    {
      "epoch": 7.0518954948521095,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 29623
    },
    {
      "epoch": 7.052133547580789,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 29624
    },
    {
      "epoch": 7.052371600309469,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 29625
    },
    {
      "epoch": 7.052609653038148,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 29626
    },
    {
      "epoch": 7.052847705766827,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 29627
    },
    {
      "epoch": 7.053085758495507,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 29628
    },
    {
      "epoch": 7.053323811224186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3088,
      "step": 29629
    },
    {
      "epoch": 7.053561863952866,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 29630
    },
    {
      "epoch": 7.053799916681545,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 29631
    },
    {
      "epoch": 7.054037969410224,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 29632
    },
    {
      "epoch": 7.054276022138904,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1191,
      "step": 29633
    },
    {
      "epoch": 7.054514074867583,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 29634
    },
    {
      "epoch": 7.054752127596262,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 29635
    },
    {
      "epoch": 7.054990180324942,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.318,
      "step": 29636
    },
    {
      "epoch": 7.055228233053621,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 29637
    },
    {
      "epoch": 7.055466285782301,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 29638
    },
    {
      "epoch": 7.0557043385109806,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 29639
    },
    {
      "epoch": 7.055942391239659,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 29640
    },
    {
      "epoch": 7.056180443968339,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 29641
    },
    {
      "epoch": 7.0564184966970185,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 29642
    },
    {
      "epoch": 7.056656549425698,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 29643
    },
    {
      "epoch": 7.056894602154377,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 29644
    },
    {
      "epoch": 7.057132654883056,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1604,
      "step": 29645
    },
    {
      "epoch": 7.057370707611736,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 29646
    },
    {
      "epoch": 7.057608760340416,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 29647
    },
    {
      "epoch": 7.057846813069095,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 29648
    },
    {
      "epoch": 7.058084865797774,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 29649
    },
    {
      "epoch": 7.0583229185264535,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 29650
    },
    {
      "epoch": 7.058560971255133,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 29651
    },
    {
      "epoch": 7.058799023983813,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 29652
    },
    {
      "epoch": 7.0590370767124915,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 29653
    },
    {
      "epoch": 7.059275129441171,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 29654
    },
    {
      "epoch": 7.059513182169851,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 29655
    },
    {
      "epoch": 7.05975123489853,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 29656
    },
    {
      "epoch": 7.059989287627209,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 29657
    },
    {
      "epoch": 7.060227340355889,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 29658
    },
    {
      "epoch": 7.060465393084568,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 29659
    },
    {
      "epoch": 7.060703445813248,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8908,
      "step": 29660
    },
    {
      "epoch": 7.060941498541927,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 29661
    },
    {
      "epoch": 7.061179551270606,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 29662
    },
    {
      "epoch": 7.061417603999286,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7713,
      "step": 29663
    },
    {
      "epoch": 7.061655656727965,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 29664
    },
    {
      "epoch": 7.061893709456645,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 29665
    },
    {
      "epoch": 7.062131762185324,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 29666
    },
    {
      "epoch": 7.062369814914003,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 29667
    },
    {
      "epoch": 7.062607867642683,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 29668
    },
    {
      "epoch": 7.0628459203713625,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 29669
    },
    {
      "epoch": 7.063083973100042,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 29670
    },
    {
      "epoch": 7.063322025828721,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 29671
    },
    {
      "epoch": 7.0635600785574,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 29672
    },
    {
      "epoch": 7.06379813128608,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3326,
      "step": 29673
    },
    {
      "epoch": 7.06403618401476,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 29674
    },
    {
      "epoch": 7.064274236743438,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 29675
    },
    {
      "epoch": 7.064512289472118,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 29676
    },
    {
      "epoch": 7.064750342200798,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 29677
    },
    {
      "epoch": 7.064988394929477,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3248,
      "step": 29678
    },
    {
      "epoch": 7.065226447658156,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 29679
    },
    {
      "epoch": 7.0654645003868355,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 29680
    },
    {
      "epoch": 7.065702553115515,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 29681
    },
    {
      "epoch": 7.065940605844195,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 29682
    },
    {
      "epoch": 7.066178658572874,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 29683
    },
    {
      "epoch": 7.066416711301553,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 29684
    },
    {
      "epoch": 7.066654764030233,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 29685
    },
    {
      "epoch": 7.066892816758912,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 29686
    },
    {
      "epoch": 7.067130869487592,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3594,
      "step": 29687
    },
    {
      "epoch": 7.0673689222162706,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 29688
    },
    {
      "epoch": 7.06760697494495,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 29689
    },
    {
      "epoch": 7.06784502767363,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1752,
      "step": 29690
    },
    {
      "epoch": 7.068083080402309,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 29691
    },
    {
      "epoch": 7.068321133130988,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 29692
    },
    {
      "epoch": 7.068559185859668,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.363,
      "step": 29693
    },
    {
      "epoch": 7.068797238588347,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 29694
    },
    {
      "epoch": 7.069035291317027,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 29695
    },
    {
      "epoch": 7.0692733440457065,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.6618,
      "step": 29696
    },
    {
      "epoch": 7.069511396774385,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1332,
      "step": 29697
    },
    {
      "epoch": 7.069749449503065,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 29698
    },
    {
      "epoch": 7.069987502231744,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 29699
    },
    {
      "epoch": 7.070225554960424,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 29700
    },
    {
      "epoch": 7.070225554960424,
      "eval_runtime": 0.5184,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 1.929,
      "step": 29700
    },
    {
      "epoch": 7.070463607689103,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3433,
      "step": 29701
    },
    {
      "epoch": 7.070701660417782,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 29702
    },
    {
      "epoch": 7.070939713146462,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 29703
    },
    {
      "epoch": 7.071177765875142,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 29704
    },
    {
      "epoch": 7.071415818603821,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3092,
      "step": 29705
    },
    {
      "epoch": 7.0716538713325,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 29706
    },
    {
      "epoch": 7.0718919240611795,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 29707
    },
    {
      "epoch": 7.072129976789859,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3316,
      "step": 29708
    },
    {
      "epoch": 7.072368029518539,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 29709
    },
    {
      "epoch": 7.072606082247217,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7032,
      "step": 29710
    },
    {
      "epoch": 7.072844134975897,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 29711
    },
    {
      "epoch": 7.073082187704577,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 29712
    },
    {
      "epoch": 7.073320240433256,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 29713
    },
    {
      "epoch": 7.073558293161935,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 29714
    },
    {
      "epoch": 7.073796345890615,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2422,
      "step": 29715
    },
    {
      "epoch": 7.074034398619294,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4233,
      "step": 29716
    },
    {
      "epoch": 7.074272451347974,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 29717
    },
    {
      "epoch": 7.074510504076653,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 29718
    },
    {
      "epoch": 7.074748556805332,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 29719
    },
    {
      "epoch": 7.074986609534012,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 29720
    },
    {
      "epoch": 7.075224662262691,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 29721
    },
    {
      "epoch": 7.075462714991371,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1142,
      "step": 29722
    },
    {
      "epoch": 7.07570076772005,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 29723
    },
    {
      "epoch": 7.075938820448729,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2866,
      "step": 29724
    },
    {
      "epoch": 7.076176873177409,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 29725
    },
    {
      "epoch": 7.0764149259060884,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 29726
    },
    {
      "epoch": 7.076652978634767,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 29727
    },
    {
      "epoch": 7.076891031363447,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 29728
    },
    {
      "epoch": 7.077129084092126,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 29729
    },
    {
      "epoch": 7.077367136820806,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 29730
    },
    {
      "epoch": 7.077605189549486,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 29731
    },
    {
      "epoch": 7.077843242278164,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 29732
    },
    {
      "epoch": 7.078081295006844,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 29733
    },
    {
      "epoch": 7.0783193477355235,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 29734
    },
    {
      "epoch": 7.078557400464203,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 29735
    },
    {
      "epoch": 7.078795453192882,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 29736
    },
    {
      "epoch": 7.079033505921561,
      "grad_norm": 7.875,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 29737
    },
    {
      "epoch": 7.079271558650241,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 29738
    },
    {
      "epoch": 7.079509611378921,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 29739
    },
    {
      "epoch": 7.0797476641076,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 29740
    },
    {
      "epoch": 7.079985716836279,
      "grad_norm": 7.53125,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 29741
    },
    {
      "epoch": 7.080223769564959,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 29742
    },
    {
      "epoch": 7.080461822293638,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 29743
    },
    {
      "epoch": 7.080699875022318,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 29744
    },
    {
      "epoch": 7.0809379277509965,
      "grad_norm": 6.4375,
      "learning_rate": 3e-05,
      "loss": 1.7827,
      "step": 29745
    },
    {
      "epoch": 7.081175980479676,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 29746
    },
    {
      "epoch": 7.081414033208356,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 29747
    },
    {
      "epoch": 7.081652085937035,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 29748
    },
    {
      "epoch": 7.081890138665714,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 29749
    },
    {
      "epoch": 7.082128191394394,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 29750
    },
    {
      "epoch": 7.082366244123073,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 29751
    },
    {
      "epoch": 7.082604296851753,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 29752
    },
    {
      "epoch": 7.0828423495804325,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.2024,
      "step": 29753
    },
    {
      "epoch": 7.083080402309111,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 29754
    },
    {
      "epoch": 7.083318455037791,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 29755
    },
    {
      "epoch": 7.08355650776647,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 29756
    },
    {
      "epoch": 7.08379456049515,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.326,
      "step": 29757
    },
    {
      "epoch": 7.084032613223829,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 29758
    },
    {
      "epoch": 7.084270665952508,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 29759
    },
    {
      "epoch": 7.084508718681188,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 29760
    },
    {
      "epoch": 7.0847467714098675,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 29761
    },
    {
      "epoch": 7.084984824138546,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 29762
    },
    {
      "epoch": 7.085222876867226,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2852,
      "step": 29763
    },
    {
      "epoch": 7.0854609295959055,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 29764
    },
    {
      "epoch": 7.085698982324585,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8031,
      "step": 29765
    },
    {
      "epoch": 7.085937035053265,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2043,
      "step": 29766
    },
    {
      "epoch": 7.086175087781943,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 29767
    },
    {
      "epoch": 7.086413140510623,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 29768
    },
    {
      "epoch": 7.086651193239303,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 29769
    },
    {
      "epoch": 7.086889245967982,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9877,
      "step": 29770
    },
    {
      "epoch": 7.087127298696661,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 29771
    },
    {
      "epoch": 7.0873653514253405,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.6887,
      "step": 29772
    },
    {
      "epoch": 7.08760340415402,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7527,
      "step": 29773
    },
    {
      "epoch": 7.0878414568827,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 29774
    },
    {
      "epoch": 7.088079509611379,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 29775
    },
    {
      "epoch": 7.088317562340058,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 29776
    },
    {
      "epoch": 7.088555615068738,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 29777
    },
    {
      "epoch": 7.088793667797417,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 29778
    },
    {
      "epoch": 7.089031720526097,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 29779
    },
    {
      "epoch": 7.089269773254776,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 29780
    },
    {
      "epoch": 7.089507825983455,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 29781
    },
    {
      "epoch": 7.089745878712135,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7842,
      "step": 29782
    },
    {
      "epoch": 7.089983931440814,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 29783
    },
    {
      "epoch": 7.090221984169493,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2469,
      "step": 29784
    },
    {
      "epoch": 7.090460036898173,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 29785
    },
    {
      "epoch": 7.090698089626852,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.882,
      "step": 29786
    },
    {
      "epoch": 7.090936142355532,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1288,
      "step": 29787
    },
    {
      "epoch": 7.0911741950842115,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 29788
    },
    {
      "epoch": 7.09141224781289,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 29789
    },
    {
      "epoch": 7.09165030054157,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 29790
    },
    {
      "epoch": 7.0918883532702495,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 29791
    },
    {
      "epoch": 7.092126405998929,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 29792
    },
    {
      "epoch": 7.092364458727608,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 29793
    },
    {
      "epoch": 7.092602511456287,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 29794
    },
    {
      "epoch": 7.092840564184967,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 29795
    },
    {
      "epoch": 7.093078616913647,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1224,
      "step": 29796
    },
    {
      "epoch": 7.093316669642326,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 29797
    },
    {
      "epoch": 7.093554722371005,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 29798
    },
    {
      "epoch": 7.0937927750996845,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.053,
      "step": 29799
    },
    {
      "epoch": 7.094030827828364,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 29800
    },
    {
      "epoch": 7.094030827828364,
      "eval_runtime": 0.5301,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 29800
    },
    {
      "epoch": 7.094268880557044,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 29801
    },
    {
      "epoch": 7.0945069332857225,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 29802
    },
    {
      "epoch": 7.094744986014402,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 29803
    },
    {
      "epoch": 7.094983038743082,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 29804
    },
    {
      "epoch": 7.095221091471761,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 29805
    },
    {
      "epoch": 7.09545914420044,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 29806
    },
    {
      "epoch": 7.09569719692912,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 29807
    },
    {
      "epoch": 7.095935249657799,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 29808
    },
    {
      "epoch": 7.096173302386479,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 29809
    },
    {
      "epoch": 7.096411355115158,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2578,
      "step": 29810
    },
    {
      "epoch": 7.096649407843837,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 29811
    },
    {
      "epoch": 7.096887460572517,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3848,
      "step": 29812
    },
    {
      "epoch": 7.097125513301196,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 29813
    },
    {
      "epoch": 7.097363566029876,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6876,
      "step": 29814
    },
    {
      "epoch": 7.097601618758555,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 29815
    },
    {
      "epoch": 7.097839671487234,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 29816
    },
    {
      "epoch": 7.098077724215914,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 29817
    },
    {
      "epoch": 7.0983157769445935,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9007,
      "step": 29818
    },
    {
      "epoch": 7.098553829673272,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 29819
    },
    {
      "epoch": 7.098791882401952,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 29820
    },
    {
      "epoch": 7.099029935130631,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9355,
      "step": 29821
    },
    {
      "epoch": 7.099267987859311,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 29822
    },
    {
      "epoch": 7.099506040587991,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 29823
    },
    {
      "epoch": 7.099744093316669,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 29824
    },
    {
      "epoch": 7.099982146045349,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 29825
    },
    {
      "epoch": 7.1002201987740285,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7446,
      "step": 29826
    },
    {
      "epoch": 7.100458251502708,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 29827
    },
    {
      "epoch": 7.100696304231387,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 29828
    },
    {
      "epoch": 7.1009343569600665,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.4287,
      "step": 29829
    },
    {
      "epoch": 7.101172409688746,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0089,
      "step": 29830
    },
    {
      "epoch": 7.101410462417426,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 29831
    },
    {
      "epoch": 7.101648515146105,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 29832
    },
    {
      "epoch": 7.101886567874784,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 29833
    },
    {
      "epoch": 7.102124620603464,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 29834
    },
    {
      "epoch": 7.102362673332143,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 29835
    },
    {
      "epoch": 7.102600726060823,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 29836
    },
    {
      "epoch": 7.1028387787895015,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 29837
    },
    {
      "epoch": 7.103076831518181,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0659,
      "step": 29838
    },
    {
      "epoch": 7.103314884246861,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 29839
    },
    {
      "epoch": 7.10355293697554,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2586,
      "step": 29840
    },
    {
      "epoch": 7.103790989704219,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 29841
    },
    {
      "epoch": 7.104029042432899,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 29842
    },
    {
      "epoch": 7.104267095161578,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 29843
    },
    {
      "epoch": 7.104505147890258,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2307,
      "step": 29844
    },
    {
      "epoch": 7.1047432006189375,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8262,
      "step": 29845
    },
    {
      "epoch": 7.104981253347616,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 29846
    },
    {
      "epoch": 7.105219306076296,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 29847
    },
    {
      "epoch": 7.105457358804975,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 29848
    },
    {
      "epoch": 7.105695411533655,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 29849
    },
    {
      "epoch": 7.105933464262334,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 29850
    },
    {
      "epoch": 7.106171516991013,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 29851
    },
    {
      "epoch": 7.106409569719693,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 29852
    },
    {
      "epoch": 7.106647622448373,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 29853
    },
    {
      "epoch": 7.106885675177052,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 29854
    },
    {
      "epoch": 7.107123727905731,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 29855
    },
    {
      "epoch": 7.1073617806344105,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 29856
    },
    {
      "epoch": 7.10759983336309,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.489,
      "step": 29857
    },
    {
      "epoch": 7.10783788609177,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0136,
      "step": 29858
    },
    {
      "epoch": 7.108075938820448,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 29859
    },
    {
      "epoch": 7.108313991549128,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 29860
    },
    {
      "epoch": 7.108552044277808,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 29861
    },
    {
      "epoch": 7.108790097006487,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9938,
      "step": 29862
    },
    {
      "epoch": 7.109028149735166,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 29863
    },
    {
      "epoch": 7.1092662024638456,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0856,
      "step": 29864
    },
    {
      "epoch": 7.109504255192525,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 29865
    },
    {
      "epoch": 7.109742307921205,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.5843,
      "step": 29866
    },
    {
      "epoch": 7.109980360649884,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 29867
    },
    {
      "epoch": 7.110218413378563,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8074,
      "step": 29868
    },
    {
      "epoch": 7.110456466107243,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 29869
    },
    {
      "epoch": 7.110694518835922,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 29870
    },
    {
      "epoch": 7.110932571564602,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 29871
    },
    {
      "epoch": 7.111170624293281,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 29872
    },
    {
      "epoch": 7.11140867702196,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 29873
    },
    {
      "epoch": 7.11164672975064,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 29874
    },
    {
      "epoch": 7.111884782479319,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 29875
    },
    {
      "epoch": 7.112122835207998,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 29876
    },
    {
      "epoch": 7.112360887936678,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 29877
    },
    {
      "epoch": 7.112598940665357,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 29878
    },
    {
      "epoch": 7.112836993394037,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 29879
    },
    {
      "epoch": 7.113075046122717,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 29880
    },
    {
      "epoch": 7.113313098851395,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 29881
    },
    {
      "epoch": 7.113551151580075,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 29882
    },
    {
      "epoch": 7.1137892043087545,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 29883
    },
    {
      "epoch": 7.114027257037434,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 29884
    },
    {
      "epoch": 7.114265309766113,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 29885
    },
    {
      "epoch": 7.114503362494792,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 29886
    },
    {
      "epoch": 7.114741415223472,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 29887
    },
    {
      "epoch": 7.114979467952152,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 29888
    },
    {
      "epoch": 7.115217520680831,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.3737,
      "step": 29889
    },
    {
      "epoch": 7.11545557340951,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 29890
    },
    {
      "epoch": 7.11569362613819,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9881,
      "step": 29891
    },
    {
      "epoch": 7.115931678866869,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 29892
    },
    {
      "epoch": 7.116169731595549,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 29893
    },
    {
      "epoch": 7.1164077843242275,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 29894
    },
    {
      "epoch": 7.116645837052907,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 29895
    },
    {
      "epoch": 7.116883889781587,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 29896
    },
    {
      "epoch": 7.117121942510266,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 29897
    },
    {
      "epoch": 7.117359995238945,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 29898
    },
    {
      "epoch": 7.117598047967625,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 29899
    },
    {
      "epoch": 7.117836100696304,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 29900
    },
    {
      "epoch": 7.117836100696304,
      "eval_runtime": 0.5251,
      "eval_samples_per_second": 1.904,
      "eval_steps_per_second": 1.904,
      "step": 29900
    },
    {
      "epoch": 7.118074153424984,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 29901
    },
    {
      "epoch": 7.1183122061536634,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0041,
      "step": 29902
    },
    {
      "epoch": 7.118550258882342,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 29903
    },
    {
      "epoch": 7.118788311611022,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 29904
    },
    {
      "epoch": 7.119026364339701,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 29905
    },
    {
      "epoch": 7.119264417068381,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 29906
    },
    {
      "epoch": 7.11950246979706,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 29907
    },
    {
      "epoch": 7.119740522525739,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 29908
    },
    {
      "epoch": 7.119978575254419,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 29909
    },
    {
      "epoch": 7.1202166279830985,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 29910
    },
    {
      "epoch": 7.120454680711777,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 29911
    },
    {
      "epoch": 7.120692733440457,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2619,
      "step": 29912
    },
    {
      "epoch": 7.120930786169136,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2557,
      "step": 29913
    },
    {
      "epoch": 7.121168838897816,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 29914
    },
    {
      "epoch": 7.121406891626496,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 29915
    },
    {
      "epoch": 7.121644944355174,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 29916
    },
    {
      "epoch": 7.121882997083854,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 29917
    },
    {
      "epoch": 7.122121049812534,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8966,
      "step": 29918
    },
    {
      "epoch": 7.122359102541213,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 29919
    },
    {
      "epoch": 7.122597155269892,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 29920
    },
    {
      "epoch": 7.1228352079985715,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 29921
    },
    {
      "epoch": 7.123073260727251,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 29922
    },
    {
      "epoch": 7.123311313455931,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 29923
    },
    {
      "epoch": 7.12354936618461,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 29924
    },
    {
      "epoch": 7.123787418913289,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 29925
    },
    {
      "epoch": 7.124025471641969,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 29926
    },
    {
      "epoch": 7.124263524370648,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 29927
    },
    {
      "epoch": 7.124501577099328,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 29928
    },
    {
      "epoch": 7.124739629828007,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3719,
      "step": 29929
    },
    {
      "epoch": 7.124977682556686,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9898,
      "step": 29930
    },
    {
      "epoch": 7.125215735285366,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 29931
    },
    {
      "epoch": 7.125453788014045,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8307,
      "step": 29932
    },
    {
      "epoch": 7.125691840742724,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 29933
    },
    {
      "epoch": 7.125929893471404,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 29934
    },
    {
      "epoch": 7.126167946200083,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7714,
      "step": 29935
    },
    {
      "epoch": 7.126405998928763,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 29936
    },
    {
      "epoch": 7.1266440516574425,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 29937
    },
    {
      "epoch": 7.126882104386121,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3001,
      "step": 29938
    },
    {
      "epoch": 7.127120157114801,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9227,
      "step": 29939
    },
    {
      "epoch": 7.1273582098434805,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 29940
    },
    {
      "epoch": 7.12759626257216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3966,
      "step": 29941
    },
    {
      "epoch": 7.127834315300839,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 29942
    },
    {
      "epoch": 7.128072368029518,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 29943
    },
    {
      "epoch": 7.128310420758198,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 29944
    },
    {
      "epoch": 7.128548473486878,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 29945
    },
    {
      "epoch": 7.128786526215556,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 29946
    },
    {
      "epoch": 7.129024578944236,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 29947
    },
    {
      "epoch": 7.1292626316729155,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.6513,
      "step": 29948
    },
    {
      "epoch": 7.129500684401595,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 29949
    },
    {
      "epoch": 7.129738737130275,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 29950
    },
    {
      "epoch": 7.1299767898589534,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8748,
      "step": 29951
    },
    {
      "epoch": 7.130214842587633,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 29952
    },
    {
      "epoch": 7.130452895316313,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 29953
    },
    {
      "epoch": 7.130690948044992,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.007,
      "step": 29954
    },
    {
      "epoch": 7.130929000773671,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3346,
      "step": 29955
    },
    {
      "epoch": 7.131167053502351,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 29956
    },
    {
      "epoch": 7.13140510623103,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 29957
    },
    {
      "epoch": 7.13164315895971,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 29958
    },
    {
      "epoch": 7.131881211688389,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 29959
    },
    {
      "epoch": 7.132119264417068,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3996,
      "step": 29960
    },
    {
      "epoch": 7.132357317145748,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 29961
    },
    {
      "epoch": 7.132595369874427,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 29962
    },
    {
      "epoch": 7.132833422603107,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 29963
    },
    {
      "epoch": 7.133071475331786,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 29964
    },
    {
      "epoch": 7.133309528060465,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 29965
    },
    {
      "epoch": 7.133547580789145,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 29966
    },
    {
      "epoch": 7.1337856335178245,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 29967
    },
    {
      "epoch": 7.134023686246503,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 29968
    },
    {
      "epoch": 7.134261738975183,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 29969
    },
    {
      "epoch": 7.134499791703862,
      "grad_norm": 8.8125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 29970
    },
    {
      "epoch": 7.134737844432542,
      "grad_norm": 7.5,
      "learning_rate": 3e-05,
      "loss": 2.1283,
      "step": 29971
    },
    {
      "epoch": 7.134975897161222,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 29972
    },
    {
      "epoch": 7.1352139498899,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 29973
    },
    {
      "epoch": 7.13545200261858,
      "grad_norm": 9.0,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 29974
    },
    {
      "epoch": 7.1356900553472595,
      "grad_norm": 8.375,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 29975
    },
    {
      "epoch": 7.135928108075939,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 29976
    },
    {
      "epoch": 7.136166160804618,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 29977
    },
    {
      "epoch": 7.1364042135332975,
      "grad_norm": 7.5,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 29978
    },
    {
      "epoch": 7.136642266261977,
      "grad_norm": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 29979
    },
    {
      "epoch": 7.136880318990657,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 29980
    },
    {
      "epoch": 7.137118371719335,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 29981
    },
    {
      "epoch": 7.137356424448015,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 29982
    },
    {
      "epoch": 7.137594477176695,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 29983
    },
    {
      "epoch": 7.137832529905374,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 29984
    },
    {
      "epoch": 7.138070582634054,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2756,
      "step": 29985
    },
    {
      "epoch": 7.1383086353627325,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 29986
    },
    {
      "epoch": 7.138546688091412,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 29987
    },
    {
      "epoch": 7.138784740820092,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 29988
    },
    {
      "epoch": 7.139022793548771,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 29989
    },
    {
      "epoch": 7.13926084627745,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 29990
    },
    {
      "epoch": 7.13949889900613,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 29991
    },
    {
      "epoch": 7.139736951734809,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8341,
      "step": 29992
    },
    {
      "epoch": 7.139975004463489,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 29993
    },
    {
      "epoch": 7.1402130571921685,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.267,
      "step": 29994
    },
    {
      "epoch": 7.140451109920847,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 29995
    },
    {
      "epoch": 7.140689162649527,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 29996
    },
    {
      "epoch": 7.140927215378206,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 29997
    },
    {
      "epoch": 7.141165268106886,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 29998
    },
    {
      "epoch": 7.141403320835565,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 29999
    },
    {
      "epoch": 7.141641373564244,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 30000
    },
    {
      "epoch": 7.141641373564244,
      "eval_runtime": 0.5037,
      "eval_samples_per_second": 1.985,
      "eval_steps_per_second": 1.985,
      "step": 30000
    },
    {
      "epoch": 7.141879426292924,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 30001
    },
    {
      "epoch": 7.1421174790216035,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 30002
    },
    {
      "epoch": 7.142355531750282,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 30003
    },
    {
      "epoch": 7.142593584478962,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 30004
    },
    {
      "epoch": 7.1428316372076415,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 30005
    },
    {
      "epoch": 7.143069689936321,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 30006
    },
    {
      "epoch": 7.143307742665001,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1769,
      "step": 30007
    },
    {
      "epoch": 7.143545795393679,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 30008
    },
    {
      "epoch": 7.143783848122359,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 30009
    },
    {
      "epoch": 7.144021900851039,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 30010
    },
    {
      "epoch": 7.144259953579718,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 30011
    },
    {
      "epoch": 7.144498006308397,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 30012
    },
    {
      "epoch": 7.1447360590370765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2383,
      "step": 30013
    },
    {
      "epoch": 7.144974111765756,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 30014
    },
    {
      "epoch": 7.145212164494436,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 30015
    },
    {
      "epoch": 7.1454502172231145,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5676,
      "step": 30016
    },
    {
      "epoch": 7.145688269951794,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 30017
    },
    {
      "epoch": 7.145926322680474,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 30018
    },
    {
      "epoch": 7.146164375409153,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2506,
      "step": 30019
    },
    {
      "epoch": 7.146402428137833,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2788,
      "step": 30020
    },
    {
      "epoch": 7.146640480866512,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 30021
    },
    {
      "epoch": 7.146878533595191,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.08,
      "step": 30022
    },
    {
      "epoch": 7.147116586323871,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 30023
    },
    {
      "epoch": 7.14735463905255,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 30024
    },
    {
      "epoch": 7.147592691781229,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 30025
    },
    {
      "epoch": 7.147830744509909,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 30026
    },
    {
      "epoch": 7.148068797238588,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 30027
    },
    {
      "epoch": 7.148306849967268,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 30028
    },
    {
      "epoch": 7.148544902695948,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 30029
    },
    {
      "epoch": 7.148782955424626,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 30030
    },
    {
      "epoch": 7.149021008153306,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2877,
      "step": 30031
    },
    {
      "epoch": 7.1492590608819855,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8395,
      "step": 30032
    },
    {
      "epoch": 7.149497113610665,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1683,
      "step": 30033
    },
    {
      "epoch": 7.149735166339344,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 30034
    },
    {
      "epoch": 7.149973219068023,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 30035
    },
    {
      "epoch": 7.150211271796703,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 30036
    },
    {
      "epoch": 7.150449324525383,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3533,
      "step": 30037
    },
    {
      "epoch": 7.150687377254062,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2266,
      "step": 30038
    },
    {
      "epoch": 7.150925429982741,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 30039
    },
    {
      "epoch": 7.1511634827114205,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.177,
      "step": 30040
    },
    {
      "epoch": 7.1514015354401,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 30041
    },
    {
      "epoch": 7.15163958816878,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 30042
    },
    {
      "epoch": 7.1518776408974585,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 30043
    },
    {
      "epoch": 7.152115693626138,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 30044
    },
    {
      "epoch": 7.152353746354818,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 30045
    },
    {
      "epoch": 7.152591799083497,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 30046
    },
    {
      "epoch": 7.152829851812176,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 30047
    },
    {
      "epoch": 7.153067904540856,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 30048
    },
    {
      "epoch": 7.153305957269535,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 30049
    },
    {
      "epoch": 7.153544009998215,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 30050
    },
    {
      "epoch": 7.153782062726894,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 30051
    },
    {
      "epoch": 7.154020115455573,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 30052
    },
    {
      "epoch": 7.154258168184253,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.339,
      "step": 30053
    },
    {
      "epoch": 7.154496220912932,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 30054
    },
    {
      "epoch": 7.154734273641612,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 30055
    },
    {
      "epoch": 7.154972326370291,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 30056
    },
    {
      "epoch": 7.15521037909897,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.4032,
      "step": 30057
    },
    {
      "epoch": 7.15544843182765,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 30058
    },
    {
      "epoch": 7.1556864845563295,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 30059
    },
    {
      "epoch": 7.155924537285008,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 30060
    },
    {
      "epoch": 7.156162590013688,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.669,
      "step": 30061
    },
    {
      "epoch": 7.156400642742367,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 30062
    },
    {
      "epoch": 7.156638695471047,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 30063
    },
    {
      "epoch": 7.156876748199727,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 30064
    },
    {
      "epoch": 7.157114800928405,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 30065
    },
    {
      "epoch": 7.157352853657085,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2936,
      "step": 30066
    },
    {
      "epoch": 7.157590906385765,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 30067
    },
    {
      "epoch": 7.157828959114444,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 30068
    },
    {
      "epoch": 7.158067011843123,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2825,
      "step": 30069
    },
    {
      "epoch": 7.1583050645718025,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 30070
    },
    {
      "epoch": 7.158543117300482,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 30071
    },
    {
      "epoch": 7.158781170029162,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 30072
    },
    {
      "epoch": 7.159019222757841,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 30073
    },
    {
      "epoch": 7.15925727548652,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1376,
      "step": 30074
    },
    {
      "epoch": 7.1594953282152,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 30075
    },
    {
      "epoch": 7.159733380943879,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 30076
    },
    {
      "epoch": 7.159971433672559,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 30077
    },
    {
      "epoch": 7.1602094864012376,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 30078
    },
    {
      "epoch": 7.160447539129917,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0446,
      "step": 30079
    },
    {
      "epoch": 7.160685591858597,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 30080
    },
    {
      "epoch": 7.160923644587276,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 30081
    },
    {
      "epoch": 7.161161697315955,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0704,
      "step": 30082
    },
    {
      "epoch": 7.161399750044635,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8502,
      "step": 30083
    },
    {
      "epoch": 7.161637802773314,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 30084
    },
    {
      "epoch": 7.161875855501994,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 30085
    },
    {
      "epoch": 7.1621139082306735,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 30086
    },
    {
      "epoch": 7.162351960959352,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 30087
    },
    {
      "epoch": 7.162590013688032,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 30088
    },
    {
      "epoch": 7.162828066416711,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 30089
    },
    {
      "epoch": 7.163066119145391,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 30090
    },
    {
      "epoch": 7.16330417187407,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 30091
    },
    {
      "epoch": 7.163542224602749,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4161,
      "step": 30092
    },
    {
      "epoch": 7.163780277331429,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 30093
    },
    {
      "epoch": 7.164018330060109,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 30094
    },
    {
      "epoch": 7.164256382788787,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 30095
    },
    {
      "epoch": 7.164494435517467,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 30096
    },
    {
      "epoch": 7.1647324882461465,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 30097
    },
    {
      "epoch": 7.164970540974826,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 30098
    },
    {
      "epoch": 7.165208593703506,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 30099
    },
    {
      "epoch": 7.165446646432184,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 30100
    },
    {
      "epoch": 7.165446646432184,
      "eval_runtime": 0.5188,
      "eval_samples_per_second": 1.928,
      "eval_steps_per_second": 1.928,
      "step": 30100
    },
    {
      "epoch": 7.165684699160864,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2405,
      "step": 30101
    },
    {
      "epoch": 7.165922751889544,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 30102
    },
    {
      "epoch": 7.166160804618223,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7233,
      "step": 30103
    },
    {
      "epoch": 7.166398857346902,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 30104
    },
    {
      "epoch": 7.166636910075582,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 30105
    },
    {
      "epoch": 7.166874962804261,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 30106
    },
    {
      "epoch": 7.167113015532941,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 30107
    },
    {
      "epoch": 7.16735106826162,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 30108
    },
    {
      "epoch": 7.167589120990299,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0627,
      "step": 30109
    },
    {
      "epoch": 7.167827173718979,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3542,
      "step": 30110
    },
    {
      "epoch": 7.168065226447658,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3819,
      "step": 30111
    },
    {
      "epoch": 7.168303279176338,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 30112
    },
    {
      "epoch": 7.168541331905017,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2471,
      "step": 30113
    },
    {
      "epoch": 7.168779384633696,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 30114
    },
    {
      "epoch": 7.169017437362376,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 30115
    },
    {
      "epoch": 7.1692554900910554,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 30116
    },
    {
      "epoch": 7.169493542819734,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 30117
    },
    {
      "epoch": 7.169731595548414,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 30118
    },
    {
      "epoch": 7.169969648277093,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 30119
    },
    {
      "epoch": 7.170207701005773,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 30120
    },
    {
      "epoch": 7.170445753734453,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 30121
    },
    {
      "epoch": 7.170683806463131,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 30122
    },
    {
      "epoch": 7.170921859191811,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 30123
    },
    {
      "epoch": 7.1711599119204905,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 30124
    },
    {
      "epoch": 7.17139796464917,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2888,
      "step": 30125
    },
    {
      "epoch": 7.171636017377849,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 30126
    },
    {
      "epoch": 7.171874070106528,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8751,
      "step": 30127
    },
    {
      "epoch": 7.172112122835208,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 30128
    },
    {
      "epoch": 7.172350175563888,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 30129
    },
    {
      "epoch": 7.172588228292566,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 30130
    },
    {
      "epoch": 7.172826281021246,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 30131
    },
    {
      "epoch": 7.173064333749926,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3931,
      "step": 30132
    },
    {
      "epoch": 7.173302386478605,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 30133
    },
    {
      "epoch": 7.173540439207285,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 30134
    },
    {
      "epoch": 7.1737784919359635,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 30135
    },
    {
      "epoch": 7.174016544664643,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 30136
    },
    {
      "epoch": 7.174254597393323,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 30137
    },
    {
      "epoch": 7.174492650122002,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 30138
    },
    {
      "epoch": 7.174730702850681,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 30139
    },
    {
      "epoch": 7.174968755579361,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 30140
    },
    {
      "epoch": 7.17520680830804,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 30141
    },
    {
      "epoch": 7.17544486103672,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 30142
    },
    {
      "epoch": 7.1756829137653995,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 30143
    },
    {
      "epoch": 7.175920966494078,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 30144
    },
    {
      "epoch": 7.176159019222758,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 30145
    },
    {
      "epoch": 7.176397071951437,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.4388,
      "step": 30146
    },
    {
      "epoch": 7.176635124680117,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 30147
    },
    {
      "epoch": 7.176873177408796,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8703,
      "step": 30148
    },
    {
      "epoch": 7.177111230137475,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 30149
    },
    {
      "epoch": 7.177349282866155,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 30150
    },
    {
      "epoch": 7.1775873355948345,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 30151
    },
    {
      "epoch": 7.177825388323513,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8275,
      "step": 30152
    },
    {
      "epoch": 7.178063441052193,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 30153
    },
    {
      "epoch": 7.1783014937808725,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.7247,
      "step": 30154
    },
    {
      "epoch": 7.178539546509552,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 30155
    },
    {
      "epoch": 7.178777599238232,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 30156
    },
    {
      "epoch": 7.17901565196691,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 30157
    },
    {
      "epoch": 7.17925370469559,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 30158
    },
    {
      "epoch": 7.17949175742427,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 30159
    },
    {
      "epoch": 7.179729810152949,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 30160
    },
    {
      "epoch": 7.179967862881628,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 30161
    },
    {
      "epoch": 7.1802059156103075,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 30162
    },
    {
      "epoch": 7.180443968338987,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 30163
    },
    {
      "epoch": 7.180682021067667,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 30164
    },
    {
      "epoch": 7.1809200737963454,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 30165
    },
    {
      "epoch": 7.181158126525025,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 30166
    },
    {
      "epoch": 7.181396179253705,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 30167
    },
    {
      "epoch": 7.181634231982384,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 30168
    },
    {
      "epoch": 7.181872284711064,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.753,
      "step": 30169
    },
    {
      "epoch": 7.182110337439743,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 30170
    },
    {
      "epoch": 7.182348390168422,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 30171
    },
    {
      "epoch": 7.182586442897102,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 30172
    },
    {
      "epoch": 7.182824495625781,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2907,
      "step": 30173
    },
    {
      "epoch": 7.18306254835446,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 30174
    },
    {
      "epoch": 7.18330060108314,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 30175
    },
    {
      "epoch": 7.183538653811819,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1255,
      "step": 30176
    },
    {
      "epoch": 7.183776706540499,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 30177
    },
    {
      "epoch": 7.1840147592691785,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 30178
    },
    {
      "epoch": 7.184252811997857,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 30179
    },
    {
      "epoch": 7.184490864726537,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 30180
    },
    {
      "epoch": 7.1847289174552165,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 30181
    },
    {
      "epoch": 7.184966970183896,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 30182
    },
    {
      "epoch": 7.185205022912575,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 30183
    },
    {
      "epoch": 7.185443075641254,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 30184
    },
    {
      "epoch": 7.185681128369934,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8928,
      "step": 30185
    },
    {
      "epoch": 7.185919181098614,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 30186
    },
    {
      "epoch": 7.186157233827292,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 30187
    },
    {
      "epoch": 7.186395286555972,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 30188
    },
    {
      "epoch": 7.1866333392846515,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 30189
    },
    {
      "epoch": 7.186871392013331,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 30190
    },
    {
      "epoch": 7.187109444742011,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 30191
    },
    {
      "epoch": 7.1873474974706895,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 30192
    },
    {
      "epoch": 7.187585550199369,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 30193
    },
    {
      "epoch": 7.187823602928049,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 30194
    },
    {
      "epoch": 7.188061655656728,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2234,
      "step": 30195
    },
    {
      "epoch": 7.188299708385407,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2295,
      "step": 30196
    },
    {
      "epoch": 7.188537761114087,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8433,
      "step": 30197
    },
    {
      "epoch": 7.188775813842766,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 30198
    },
    {
      "epoch": 7.189013866571446,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0832,
      "step": 30199
    },
    {
      "epoch": 7.1892519193001245,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 30200
    },
    {
      "epoch": 7.1892519193001245,
      "eval_runtime": 0.524,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 30200
    },
    {
      "epoch": 7.189489972028804,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 30201
    },
    {
      "epoch": 7.189728024757484,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 30202
    },
    {
      "epoch": 7.189966077486163,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 30203
    },
    {
      "epoch": 7.190204130214843,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7718,
      "step": 30204
    },
    {
      "epoch": 7.190442182943522,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 30205
    },
    {
      "epoch": 7.190680235672201,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 30206
    },
    {
      "epoch": 7.190918288400881,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 30207
    },
    {
      "epoch": 7.1911563411295605,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.982,
      "step": 30208
    },
    {
      "epoch": 7.191394393858239,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 2.0085,
      "step": 30209
    },
    {
      "epoch": 7.191632446586919,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 30210
    },
    {
      "epoch": 7.191870499315598,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3466,
      "step": 30211
    },
    {
      "epoch": 7.192108552044278,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 30212
    },
    {
      "epoch": 7.192346604772958,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 30213
    },
    {
      "epoch": 7.192584657501636,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 30214
    },
    {
      "epoch": 7.192822710230316,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.5313,
      "step": 30215
    },
    {
      "epoch": 7.1930607629589955,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 30216
    },
    {
      "epoch": 7.193298815687675,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 30217
    },
    {
      "epoch": 7.193536868416354,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 30218
    },
    {
      "epoch": 7.1937749211450335,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 30219
    },
    {
      "epoch": 7.194012973873713,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.6,
      "step": 30220
    },
    {
      "epoch": 7.194251026602393,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7755,
      "step": 30221
    },
    {
      "epoch": 7.194489079331072,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 30222
    },
    {
      "epoch": 7.194727132059751,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 30223
    },
    {
      "epoch": 7.194965184788431,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 30224
    },
    {
      "epoch": 7.19520323751711,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 30225
    },
    {
      "epoch": 7.19544129024579,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 30226
    },
    {
      "epoch": 7.1956793429744685,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8134,
      "step": 30227
    },
    {
      "epoch": 7.195917395703148,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 30228
    },
    {
      "epoch": 7.196155448431828,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 30229
    },
    {
      "epoch": 7.196393501160507,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 30230
    },
    {
      "epoch": 7.196631553889186,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 30231
    },
    {
      "epoch": 7.196869606617866,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 30232
    },
    {
      "epoch": 7.197107659346545,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 30233
    },
    {
      "epoch": 7.197345712075225,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3052,
      "step": 30234
    },
    {
      "epoch": 7.1975837648039045,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 30235
    },
    {
      "epoch": 7.197821817532583,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8317,
      "step": 30236
    },
    {
      "epoch": 7.198059870261263,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8443,
      "step": 30237
    },
    {
      "epoch": 7.198297922989942,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 30238
    },
    {
      "epoch": 7.198535975718622,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 30239
    },
    {
      "epoch": 7.198774028447301,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 30240
    },
    {
      "epoch": 7.19901208117598,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 30241
    },
    {
      "epoch": 7.19925013390466,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 30242
    },
    {
      "epoch": 7.19948818663334,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 30243
    },
    {
      "epoch": 7.199726239362018,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 30244
    },
    {
      "epoch": 7.199964292090698,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 30245
    },
    {
      "epoch": 7.2002023448193775,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 30246
    },
    {
      "epoch": 7.200440397548057,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 30247
    },
    {
      "epoch": 7.200678450276737,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 30248
    },
    {
      "epoch": 7.200916503005415,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 30249
    },
    {
      "epoch": 7.201154555734095,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2219,
      "step": 30250
    },
    {
      "epoch": 7.201392608462775,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 30251
    },
    {
      "epoch": 7.201630661191454,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 30252
    },
    {
      "epoch": 7.201868713920133,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 30253
    },
    {
      "epoch": 7.2021067666488126,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 30254
    },
    {
      "epoch": 7.202344819377492,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0073,
      "step": 30255
    },
    {
      "epoch": 7.202582872106172,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 30256
    },
    {
      "epoch": 7.202820924834851,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 30257
    },
    {
      "epoch": 7.20305897756353,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 30258
    },
    {
      "epoch": 7.20329703029221,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7627,
      "step": 30259
    },
    {
      "epoch": 7.203535083020889,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3058,
      "step": 30260
    },
    {
      "epoch": 7.203773135749569,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 30261
    },
    {
      "epoch": 7.204011188478248,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 30262
    },
    {
      "epoch": 7.204249241206927,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6266,
      "step": 30263
    },
    {
      "epoch": 7.204487293935607,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 30264
    },
    {
      "epoch": 7.204725346664286,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 30265
    },
    {
      "epoch": 7.204963399392965,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 30266
    },
    {
      "epoch": 7.205201452121645,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7304,
      "step": 30267
    },
    {
      "epoch": 7.205439504850324,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 30268
    },
    {
      "epoch": 7.205677557579004,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 30269
    },
    {
      "epoch": 7.205915610307684,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2847,
      "step": 30270
    },
    {
      "epoch": 7.206153663036362,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 30271
    },
    {
      "epoch": 7.206391715765042,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 30272
    },
    {
      "epoch": 7.2066297684937215,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 30273
    },
    {
      "epoch": 7.206867821222401,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.296,
      "step": 30274
    },
    {
      "epoch": 7.20710587395108,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 30275
    },
    {
      "epoch": 7.207343926679759,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 30276
    },
    {
      "epoch": 7.207581979408439,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3154,
      "step": 30277
    },
    {
      "epoch": 7.207820032137119,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 30278
    },
    {
      "epoch": 7.208058084865797,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 30279
    },
    {
      "epoch": 7.208296137594477,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 30280
    },
    {
      "epoch": 7.208534190323157,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 30281
    },
    {
      "epoch": 7.208772243051836,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9403,
      "step": 30282
    },
    {
      "epoch": 7.209010295780516,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 30283
    },
    {
      "epoch": 7.2092483485091945,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 30284
    },
    {
      "epoch": 7.209486401237874,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 30285
    },
    {
      "epoch": 7.209724453966554,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 30286
    },
    {
      "epoch": 7.209962506695233,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 30287
    },
    {
      "epoch": 7.210200559423912,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 30288
    },
    {
      "epoch": 7.210438612152592,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 30289
    },
    {
      "epoch": 7.210676664881271,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 30290
    },
    {
      "epoch": 7.210914717609951,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 30291
    },
    {
      "epoch": 7.2111527703386304,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 30292
    },
    {
      "epoch": 7.211390823067309,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 30293
    },
    {
      "epoch": 7.211628875795989,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 30294
    },
    {
      "epoch": 7.211866928524668,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 30295
    },
    {
      "epoch": 7.212104981253348,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 30296
    },
    {
      "epoch": 7.212343033982027,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1368,
      "step": 30297
    },
    {
      "epoch": 7.212581086710706,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.943,
      "step": 30298
    },
    {
      "epoch": 7.212819139439386,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2982,
      "step": 30299
    },
    {
      "epoch": 7.2130571921680655,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8026,
      "step": 30300
    },
    {
      "epoch": 7.2130571921680655,
      "eval_runtime": 0.5414,
      "eval_samples_per_second": 1.847,
      "eval_steps_per_second": 1.847,
      "step": 30300
    },
    {
      "epoch": 7.213295244896744,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.814,
      "step": 30301
    },
    {
      "epoch": 7.213533297625424,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2575,
      "step": 30302
    },
    {
      "epoch": 7.213771350354103,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 30303
    },
    {
      "epoch": 7.214009403082783,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 30304
    },
    {
      "epoch": 7.214247455811463,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 30305
    },
    {
      "epoch": 7.214485508540141,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 30306
    },
    {
      "epoch": 7.214723561268821,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1978,
      "step": 30307
    },
    {
      "epoch": 7.214961613997501,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7416,
      "step": 30308
    },
    {
      "epoch": 7.21519966672618,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1229,
      "step": 30309
    },
    {
      "epoch": 7.215437719454859,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 30310
    },
    {
      "epoch": 7.2156757721835385,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 30311
    },
    {
      "epoch": 7.215913824912218,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 30312
    },
    {
      "epoch": 7.216151877640898,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 30313
    },
    {
      "epoch": 7.216389930369576,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 30314
    },
    {
      "epoch": 7.216627983098256,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 30315
    },
    {
      "epoch": 7.216866035826936,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 30316
    },
    {
      "epoch": 7.217104088555615,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 30317
    },
    {
      "epoch": 7.217342141284295,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 30318
    },
    {
      "epoch": 7.217580194012974,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 30319
    },
    {
      "epoch": 7.217818246741653,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 30320
    },
    {
      "epoch": 7.218056299470333,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 30321
    },
    {
      "epoch": 7.218294352199012,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 30322
    },
    {
      "epoch": 7.218532404927691,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 30323
    },
    {
      "epoch": 7.218770457656371,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 30324
    },
    {
      "epoch": 7.21900851038505,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 30325
    },
    {
      "epoch": 7.21924656311373,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 30326
    },
    {
      "epoch": 7.2194846158424095,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 30327
    },
    {
      "epoch": 7.219722668571088,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 30328
    },
    {
      "epoch": 7.219960721299768,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 30329
    },
    {
      "epoch": 7.2201987740284475,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0841,
      "step": 30330
    },
    {
      "epoch": 7.220436826757127,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9457,
      "step": 30331
    },
    {
      "epoch": 7.220674879485806,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 30332
    },
    {
      "epoch": 7.220912932214485,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 30333
    },
    {
      "epoch": 7.221150984943165,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 30334
    },
    {
      "epoch": 7.221389037671845,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 30335
    },
    {
      "epoch": 7.221627090400523,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 30336
    },
    {
      "epoch": 7.221865143129203,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 30337
    },
    {
      "epoch": 7.2221031958578825,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 30338
    },
    {
      "epoch": 7.222341248586562,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 30339
    },
    {
      "epoch": 7.222579301315242,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 30340
    },
    {
      "epoch": 7.2228173540439204,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 30341
    },
    {
      "epoch": 7.2230554067726,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 30342
    },
    {
      "epoch": 7.22329345950128,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 30343
    },
    {
      "epoch": 7.223531512229959,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 30344
    },
    {
      "epoch": 7.223769564958638,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 30345
    },
    {
      "epoch": 7.224007617687318,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7891,
      "step": 30346
    },
    {
      "epoch": 7.224245670415997,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 30347
    },
    {
      "epoch": 7.224483723144677,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 30348
    },
    {
      "epoch": 7.2247217758733555,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 30349
    },
    {
      "epoch": 7.224959828602035,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 30350
    },
    {
      "epoch": 7.225197881330715,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 30351
    },
    {
      "epoch": 7.225435934059394,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 30352
    },
    {
      "epoch": 7.225673986788074,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 30353
    },
    {
      "epoch": 7.225912039516753,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 30354
    },
    {
      "epoch": 7.226150092245432,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 30355
    },
    {
      "epoch": 7.226388144974112,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 30356
    },
    {
      "epoch": 7.2266261977027915,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.4031,
      "step": 30357
    },
    {
      "epoch": 7.22686425043147,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 30358
    },
    {
      "epoch": 7.22710230316015,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 30359
    },
    {
      "epoch": 7.227340355888829,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2108,
      "step": 30360
    },
    {
      "epoch": 7.227578408617509,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8186,
      "step": 30361
    },
    {
      "epoch": 7.227816461346189,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2773,
      "step": 30362
    },
    {
      "epoch": 7.228054514074867,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 30363
    },
    {
      "epoch": 7.228292566803547,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8483,
      "step": 30364
    },
    {
      "epoch": 7.2285306195322265,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 30365
    },
    {
      "epoch": 7.228768672260906,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3336,
      "step": 30366
    },
    {
      "epoch": 7.229006724989585,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 30367
    },
    {
      "epoch": 7.2292447777182645,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 30368
    },
    {
      "epoch": 7.229482830446944,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 30369
    },
    {
      "epoch": 7.229720883175624,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 30370
    },
    {
      "epoch": 7.229958935904302,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 30371
    },
    {
      "epoch": 7.230196988632982,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 30372
    },
    {
      "epoch": 7.230435041361662,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 30373
    },
    {
      "epoch": 7.230673094090341,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2644,
      "step": 30374
    },
    {
      "epoch": 7.230911146819021,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 30375
    },
    {
      "epoch": 7.2311491995476995,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 30376
    },
    {
      "epoch": 7.231387252276379,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 30377
    },
    {
      "epoch": 7.231625305005059,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2204,
      "step": 30378
    },
    {
      "epoch": 7.231863357733738,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 30379
    },
    {
      "epoch": 7.232101410462417,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 30380
    },
    {
      "epoch": 7.232339463191097,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8557,
      "step": 30381
    },
    {
      "epoch": 7.232577515919776,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 30382
    },
    {
      "epoch": 7.232815568648456,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 30383
    },
    {
      "epoch": 7.233053621377135,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 30384
    },
    {
      "epoch": 7.233291674105814,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 30385
    },
    {
      "epoch": 7.233529726834494,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 30386
    },
    {
      "epoch": 7.233767779563173,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 30387
    },
    {
      "epoch": 7.234005832291853,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3298,
      "step": 30388
    },
    {
      "epoch": 7.234243885020532,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7031,
      "step": 30389
    },
    {
      "epoch": 7.234481937749211,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 30390
    },
    {
      "epoch": 7.234719990477891,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 30391
    },
    {
      "epoch": 7.2349580432065705,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 30392
    },
    {
      "epoch": 7.235196095935249,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 30393
    },
    {
      "epoch": 7.235434148663929,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 30394
    },
    {
      "epoch": 7.2356722013926085,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 30395
    },
    {
      "epoch": 7.235910254121288,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 30396
    },
    {
      "epoch": 7.236148306849968,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 30397
    },
    {
      "epoch": 7.236386359578646,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 30398
    },
    {
      "epoch": 7.236624412307326,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 30399
    },
    {
      "epoch": 7.236862465036006,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 30400
    },
    {
      "epoch": 7.236862465036006,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 30400
    },
    {
      "epoch": 7.237100517764685,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 30401
    },
    {
      "epoch": 7.237338570493364,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2271,
      "step": 30402
    },
    {
      "epoch": 7.2375766232220435,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 30403
    },
    {
      "epoch": 7.237814675950723,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7601,
      "step": 30404
    },
    {
      "epoch": 7.238052728679403,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 30405
    },
    {
      "epoch": 7.2382907814080815,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 30406
    },
    {
      "epoch": 7.238528834136761,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 30407
    },
    {
      "epoch": 7.238766886865441,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2544,
      "step": 30408
    },
    {
      "epoch": 7.23900493959412,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0575,
      "step": 30409
    },
    {
      "epoch": 7.2392429923228,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 30410
    },
    {
      "epoch": 7.239481045051479,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 30411
    },
    {
      "epoch": 7.239719097780158,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2594,
      "step": 30412
    },
    {
      "epoch": 7.239957150508838,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 30413
    },
    {
      "epoch": 7.240195203237517,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 30414
    },
    {
      "epoch": 7.240433255966196,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 30415
    },
    {
      "epoch": 7.240671308694876,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 30416
    },
    {
      "epoch": 7.240909361423555,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3047,
      "step": 30417
    },
    {
      "epoch": 7.241147414152235,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 30418
    },
    {
      "epoch": 7.241385466880914,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 30419
    },
    {
      "epoch": 7.241623519609593,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 30420
    },
    {
      "epoch": 7.241861572338273,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8735,
      "step": 30421
    },
    {
      "epoch": 7.2420996250669525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 30422
    },
    {
      "epoch": 7.242337677795632,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 30423
    },
    {
      "epoch": 7.242575730524311,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2473,
      "step": 30424
    },
    {
      "epoch": 7.24281378325299,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 30425
    },
    {
      "epoch": 7.24305183598167,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 30426
    },
    {
      "epoch": 7.24328988871035,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 30427
    },
    {
      "epoch": 7.243527941439028,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8053,
      "step": 30428
    },
    {
      "epoch": 7.243765994167708,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 30429
    },
    {
      "epoch": 7.2440040468963875,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 30430
    },
    {
      "epoch": 7.244242099625067,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 30431
    },
    {
      "epoch": 7.244480152353747,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1428,
      "step": 30432
    },
    {
      "epoch": 7.2447182050824255,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 30433
    },
    {
      "epoch": 7.244956257811105,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 30434
    },
    {
      "epoch": 7.245194310539785,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 30435
    },
    {
      "epoch": 7.245432363268464,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 30436
    },
    {
      "epoch": 7.245670415997143,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 30437
    },
    {
      "epoch": 7.245908468725823,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 30438
    },
    {
      "epoch": 7.246146521454502,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 30439
    },
    {
      "epoch": 7.246384574183182,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 30440
    },
    {
      "epoch": 7.246622626911861,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 30441
    },
    {
      "epoch": 7.24686067964054,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 30442
    },
    {
      "epoch": 7.24709873236922,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1015,
      "step": 30443
    },
    {
      "epoch": 7.247336785097899,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 30444
    },
    {
      "epoch": 7.247574837826579,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 30445
    },
    {
      "epoch": 7.247812890555258,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 30446
    },
    {
      "epoch": 7.248050943283937,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 30447
    },
    {
      "epoch": 7.248288996012617,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 30448
    },
    {
      "epoch": 7.2485270487412965,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 30449
    },
    {
      "epoch": 7.248765101469975,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 30450
    },
    {
      "epoch": 7.249003154198655,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 30451
    },
    {
      "epoch": 7.249241206927334,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 30452
    },
    {
      "epoch": 7.249479259656014,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 30453
    },
    {
      "epoch": 7.249717312384694,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 30454
    },
    {
      "epoch": 7.249955365113372,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 30455
    },
    {
      "epoch": 7.250193417842052,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 30456
    },
    {
      "epoch": 7.250431470570732,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 30457
    },
    {
      "epoch": 7.250669523299411,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 30458
    },
    {
      "epoch": 7.25090757602809,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 30459
    },
    {
      "epoch": 7.2511456287567695,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 30460
    },
    {
      "epoch": 7.251383681485449,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3664,
      "step": 30461
    },
    {
      "epoch": 7.251621734214129,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 30462
    },
    {
      "epoch": 7.251859786942807,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8028,
      "step": 30463
    },
    {
      "epoch": 7.252097839671487,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 30464
    },
    {
      "epoch": 7.252335892400167,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 30465
    },
    {
      "epoch": 7.252573945128846,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 30466
    },
    {
      "epoch": 7.252811997857526,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 30467
    },
    {
      "epoch": 7.2530500505862046,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 30468
    },
    {
      "epoch": 7.253288103314884,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3382,
      "step": 30469
    },
    {
      "epoch": 7.253526156043564,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 30470
    },
    {
      "epoch": 7.253764208772243,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 30471
    },
    {
      "epoch": 7.254002261500922,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8345,
      "step": 30472
    },
    {
      "epoch": 7.254240314229602,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 30473
    },
    {
      "epoch": 7.254478366958281,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 30474
    },
    {
      "epoch": 7.254716419686961,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 30475
    },
    {
      "epoch": 7.2549544724156405,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 30476
    },
    {
      "epoch": 7.255192525144319,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8394,
      "step": 30477
    },
    {
      "epoch": 7.255430577872999,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 30478
    },
    {
      "epoch": 7.255668630601678,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 30479
    },
    {
      "epoch": 7.255906683330358,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 30480
    },
    {
      "epoch": 7.256144736059037,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 30481
    },
    {
      "epoch": 7.256382788787716,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.309,
      "step": 30482
    },
    {
      "epoch": 7.256620841516396,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 30483
    },
    {
      "epoch": 7.256858894245076,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 30484
    },
    {
      "epoch": 7.257096946973754,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 30485
    },
    {
      "epoch": 7.257334999702434,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 30486
    },
    {
      "epoch": 7.2575730524311135,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 30487
    },
    {
      "epoch": 7.257811105159793,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 30488
    },
    {
      "epoch": 7.258049157888472,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 30489
    },
    {
      "epoch": 7.258287210617151,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0005,
      "step": 30490
    },
    {
      "epoch": 7.258525263345831,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 30491
    },
    {
      "epoch": 7.258763316074511,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 30492
    },
    {
      "epoch": 7.25900136880319,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 30493
    },
    {
      "epoch": 7.259239421531869,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3033,
      "step": 30494
    },
    {
      "epoch": 7.259477474260549,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 30495
    },
    {
      "epoch": 7.259715526989228,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.5115,
      "step": 30496
    },
    {
      "epoch": 7.259953579717908,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 30497
    },
    {
      "epoch": 7.2601916324465865,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 30498
    },
    {
      "epoch": 7.260429685175266,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 30499
    },
    {
      "epoch": 7.260667737903946,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.236,
      "step": 30500
    },
    {
      "epoch": 7.260667737903946,
      "eval_runtime": 0.52,
      "eval_samples_per_second": 1.923,
      "eval_steps_per_second": 1.923,
      "step": 30500
    },
    {
      "epoch": 7.260905790632625,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 30501
    },
    {
      "epoch": 7.261143843361305,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 30502
    },
    {
      "epoch": 7.261381896089984,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1203,
      "step": 30503
    },
    {
      "epoch": 7.261619948818663,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 30504
    },
    {
      "epoch": 7.261858001547343,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 30505
    },
    {
      "epoch": 7.2620960542760224,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 30506
    },
    {
      "epoch": 7.262334107004701,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 30507
    },
    {
      "epoch": 7.262572159733381,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 30508
    },
    {
      "epoch": 7.26281021246206,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 30509
    },
    {
      "epoch": 7.26304826519074,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 30510
    },
    {
      "epoch": 7.26328631791942,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 30511
    },
    {
      "epoch": 7.263524370648098,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 30512
    },
    {
      "epoch": 7.263762423376778,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4509,
      "step": 30513
    },
    {
      "epoch": 7.2640004761054575,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 30514
    },
    {
      "epoch": 7.264238528834137,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 30515
    },
    {
      "epoch": 7.264476581562816,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 30516
    },
    {
      "epoch": 7.264714634291495,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2758,
      "step": 30517
    },
    {
      "epoch": 7.264952687020175,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 30518
    },
    {
      "epoch": 7.265190739748855,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 30519
    },
    {
      "epoch": 7.265428792477533,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.23,
      "step": 30520
    },
    {
      "epoch": 7.265666845206213,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 30521
    },
    {
      "epoch": 7.265904897934893,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 30522
    },
    {
      "epoch": 7.266142950663572,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 30523
    },
    {
      "epoch": 7.266381003392252,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 30524
    },
    {
      "epoch": 7.2666190561209305,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 30525
    },
    {
      "epoch": 7.26685710884961,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 30526
    },
    {
      "epoch": 7.26709516157829,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 30527
    },
    {
      "epoch": 7.267333214306969,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 30528
    },
    {
      "epoch": 7.267571267035648,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 30529
    },
    {
      "epoch": 7.267809319764328,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 30530
    },
    {
      "epoch": 7.268047372493007,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0629,
      "step": 30531
    },
    {
      "epoch": 7.268285425221687,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 30532
    },
    {
      "epoch": 7.268523477950366,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 30533
    },
    {
      "epoch": 7.268761530679045,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 30534
    },
    {
      "epoch": 7.268999583407725,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 30535
    },
    {
      "epoch": 7.269237636136404,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 30536
    },
    {
      "epoch": 7.269475688865084,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9454,
      "step": 30537
    },
    {
      "epoch": 7.269713741593763,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 30538
    },
    {
      "epoch": 7.269951794322442,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 30539
    },
    {
      "epoch": 7.270189847051122,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9404,
      "step": 30540
    },
    {
      "epoch": 7.2704278997798015,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4502,
      "step": 30541
    },
    {
      "epoch": 7.27066595250848,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2698,
      "step": 30542
    },
    {
      "epoch": 7.27090400523716,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 30543
    },
    {
      "epoch": 7.2711420579658395,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4677,
      "step": 30544
    },
    {
      "epoch": 7.271380110694519,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3624,
      "step": 30545
    },
    {
      "epoch": 7.271618163423199,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 30546
    },
    {
      "epoch": 7.271856216151877,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 30547
    },
    {
      "epoch": 7.272094268880557,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 30548
    },
    {
      "epoch": 7.272332321609237,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 30549
    },
    {
      "epoch": 7.272570374337916,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2571,
      "step": 30550
    },
    {
      "epoch": 7.272808427066595,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 30551
    },
    {
      "epoch": 7.2730464797952745,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 30552
    },
    {
      "epoch": 7.273284532523954,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 30553
    },
    {
      "epoch": 7.273522585252634,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 30554
    },
    {
      "epoch": 7.2737606379813124,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.3283,
      "step": 30555
    },
    {
      "epoch": 7.273998690709992,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 30556
    },
    {
      "epoch": 7.274236743438672,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 30557
    },
    {
      "epoch": 7.274474796167351,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 30558
    },
    {
      "epoch": 7.274712848896031,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 30559
    },
    {
      "epoch": 7.27495090162471,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 30560
    },
    {
      "epoch": 7.275188954353389,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 30561
    },
    {
      "epoch": 7.275427007082069,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 30562
    },
    {
      "epoch": 7.275665059810748,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 30563
    },
    {
      "epoch": 7.275903112539427,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 30564
    },
    {
      "epoch": 7.276141165268107,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 30565
    },
    {
      "epoch": 7.276379217996786,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 30566
    },
    {
      "epoch": 7.276617270725466,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 30567
    },
    {
      "epoch": 7.276855323454145,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 30568
    },
    {
      "epoch": 7.277093376182824,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.4722,
      "step": 30569
    },
    {
      "epoch": 7.277331428911504,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4746,
      "step": 30570
    },
    {
      "epoch": 7.2775694816401835,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 30571
    },
    {
      "epoch": 7.277807534368863,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 30572
    },
    {
      "epoch": 7.278045587097542,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 30573
    },
    {
      "epoch": 7.278283639826221,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 30574
    },
    {
      "epoch": 7.278521692554901,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 30575
    },
    {
      "epoch": 7.278759745283581,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 30576
    },
    {
      "epoch": 7.278997798012259,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 30577
    },
    {
      "epoch": 7.279235850740939,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 30578
    },
    {
      "epoch": 7.2794739034696185,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 30579
    },
    {
      "epoch": 7.279711956198298,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 30580
    },
    {
      "epoch": 7.279950008926978,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7775,
      "step": 30581
    },
    {
      "epoch": 7.2801880616556565,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 30582
    },
    {
      "epoch": 7.280426114384336,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9634,
      "step": 30583
    },
    {
      "epoch": 7.280664167113016,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 30584
    },
    {
      "epoch": 7.280902219841695,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 30585
    },
    {
      "epoch": 7.281140272570374,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3541,
      "step": 30586
    },
    {
      "epoch": 7.281378325299054,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 30587
    },
    {
      "epoch": 7.281616378027733,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 30588
    },
    {
      "epoch": 7.281854430756413,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 30589
    },
    {
      "epoch": 7.282092483485092,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 30590
    },
    {
      "epoch": 7.282330536213771,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 30591
    },
    {
      "epoch": 7.282568588942451,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3713,
      "step": 30592
    },
    {
      "epoch": 7.28280664167113,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 30593
    },
    {
      "epoch": 7.28304469439981,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 30594
    },
    {
      "epoch": 7.283282747128489,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 30595
    },
    {
      "epoch": 7.283520799857168,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 30596
    },
    {
      "epoch": 7.283758852585848,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 30597
    },
    {
      "epoch": 7.2839969053145275,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 30598
    },
    {
      "epoch": 7.284234958043206,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2812,
      "step": 30599
    },
    {
      "epoch": 7.284473010771886,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 30600
    },
    {
      "epoch": 7.284473010771886,
      "eval_runtime": 0.5152,
      "eval_samples_per_second": 1.941,
      "eval_steps_per_second": 1.941,
      "step": 30600
    },
    {
      "epoch": 7.284711063500565,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2253,
      "step": 30601
    },
    {
      "epoch": 7.284949116229245,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 30602
    },
    {
      "epoch": 7.285187168957924,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 30603
    },
    {
      "epoch": 7.285425221686603,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 30604
    },
    {
      "epoch": 7.285663274415283,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 30605
    },
    {
      "epoch": 7.2859013271439625,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 30606
    },
    {
      "epoch": 7.286139379872642,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 30607
    },
    {
      "epoch": 7.286377432601321,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 30608
    },
    {
      "epoch": 7.2866154853300005,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.228,
      "step": 30609
    },
    {
      "epoch": 7.28685353805868,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 30610
    },
    {
      "epoch": 7.28709159078736,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 30611
    },
    {
      "epoch": 7.287329643516038,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 30612
    },
    {
      "epoch": 7.287567696244718,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 30613
    },
    {
      "epoch": 7.287805748973398,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 30614
    },
    {
      "epoch": 7.288043801702077,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 30615
    },
    {
      "epoch": 7.288281854430757,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 30616
    },
    {
      "epoch": 7.2885199071594355,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 30617
    },
    {
      "epoch": 7.288757959888115,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 30618
    },
    {
      "epoch": 7.288996012616795,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 30619
    },
    {
      "epoch": 7.289234065345474,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 30620
    },
    {
      "epoch": 7.289472118074153,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 30621
    },
    {
      "epoch": 7.289710170802833,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 30622
    },
    {
      "epoch": 7.289948223531512,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 30623
    },
    {
      "epoch": 7.290186276260192,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 30624
    },
    {
      "epoch": 7.2904243289888715,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 30625
    },
    {
      "epoch": 7.29066238171755,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 30626
    },
    {
      "epoch": 7.29090043444623,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2595,
      "step": 30627
    },
    {
      "epoch": 7.291138487174909,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 30628
    },
    {
      "epoch": 7.291376539903589,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 30629
    },
    {
      "epoch": 7.291614592632268,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 30630
    },
    {
      "epoch": 7.291852645360947,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 30631
    },
    {
      "epoch": 7.292090698089627,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3086,
      "step": 30632
    },
    {
      "epoch": 7.292328750818307,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 30633
    },
    {
      "epoch": 7.292566803546985,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 30634
    },
    {
      "epoch": 7.292804856275665,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 30635
    },
    {
      "epoch": 7.2930429090043445,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 30636
    },
    {
      "epoch": 7.293280961733024,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7719,
      "step": 30637
    },
    {
      "epoch": 7.293519014461703,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 30638
    },
    {
      "epoch": 7.293757067190382,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 30639
    },
    {
      "epoch": 7.293995119919062,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 30640
    },
    {
      "epoch": 7.294233172647742,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1219,
      "step": 30641
    },
    {
      "epoch": 7.294471225376421,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 30642
    },
    {
      "epoch": 7.2947092781051,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 30643
    },
    {
      "epoch": 7.2949473308337796,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 30644
    },
    {
      "epoch": 7.295185383562459,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 30645
    },
    {
      "epoch": 7.295423436291139,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8661,
      "step": 30646
    },
    {
      "epoch": 7.2956614890198175,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 30647
    },
    {
      "epoch": 7.295899541748497,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 30648
    },
    {
      "epoch": 7.296137594477177,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 30649
    },
    {
      "epoch": 7.296375647205856,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 30650
    },
    {
      "epoch": 7.296613699934536,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 30651
    },
    {
      "epoch": 7.296851752663215,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7632,
      "step": 30652
    },
    {
      "epoch": 7.297089805391894,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2498,
      "step": 30653
    },
    {
      "epoch": 7.297327858120574,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 30654
    },
    {
      "epoch": 7.297565910849253,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 30655
    },
    {
      "epoch": 7.297803963577932,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 30656
    },
    {
      "epoch": 7.298042016306612,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 30657
    },
    {
      "epoch": 7.298280069035291,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 30658
    },
    {
      "epoch": 7.298518121763971,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 30659
    },
    {
      "epoch": 7.298756174492651,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 1.9414,
      "step": 30660
    },
    {
      "epoch": 7.298994227221329,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 30661
    },
    {
      "epoch": 7.299232279950009,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 30662
    },
    {
      "epoch": 7.2994703326786885,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 30663
    },
    {
      "epoch": 7.299708385407368,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 30664
    },
    {
      "epoch": 7.299946438136047,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 30665
    },
    {
      "epoch": 7.300184490864726,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 30666
    },
    {
      "epoch": 7.300422543593406,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7638,
      "step": 30667
    },
    {
      "epoch": 7.300660596322086,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 30668
    },
    {
      "epoch": 7.300898649050764,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 30669
    },
    {
      "epoch": 7.301136701779444,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 30670
    },
    {
      "epoch": 7.301374754508124,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 30671
    },
    {
      "epoch": 7.301612807236803,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 30672
    },
    {
      "epoch": 7.301850859965482,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 30673
    },
    {
      "epoch": 7.3020889126941615,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 30674
    },
    {
      "epoch": 7.302326965422841,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 30675
    },
    {
      "epoch": 7.302565018151521,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3187,
      "step": 30676
    },
    {
      "epoch": 7.3028030708802,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 30677
    },
    {
      "epoch": 7.303041123608879,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 30678
    },
    {
      "epoch": 7.303279176337559,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 30679
    },
    {
      "epoch": 7.303517229066238,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 30680
    },
    {
      "epoch": 7.303755281794918,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 30681
    },
    {
      "epoch": 7.303993334523597,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 30682
    },
    {
      "epoch": 7.304231387252276,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 30683
    },
    {
      "epoch": 7.304469439980956,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 30684
    },
    {
      "epoch": 7.304707492709635,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 30685
    },
    {
      "epoch": 7.304945545438315,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 30686
    },
    {
      "epoch": 7.305183598166994,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 30687
    },
    {
      "epoch": 7.305421650895673,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8568,
      "step": 30688
    },
    {
      "epoch": 7.305659703624353,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 30689
    },
    {
      "epoch": 7.3058977563530325,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 30690
    },
    {
      "epoch": 7.306135809081711,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 30691
    },
    {
      "epoch": 7.306373861810391,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 30692
    },
    {
      "epoch": 7.30661191453907,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 30693
    },
    {
      "epoch": 7.30684996726775,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 30694
    },
    {
      "epoch": 7.30708801999643,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1157,
      "step": 30695
    },
    {
      "epoch": 7.307326072725108,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8807,
      "step": 30696
    },
    {
      "epoch": 7.307564125453788,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8192,
      "step": 30697
    },
    {
      "epoch": 7.307802178182468,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 30698
    },
    {
      "epoch": 7.308040230911147,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 30699
    },
    {
      "epoch": 7.308278283639826,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 30700
    },
    {
      "epoch": 7.308278283639826,
      "eval_runtime": 0.5182,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 30700
    },
    {
      "epoch": 7.3085163363685055,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 30701
    },
    {
      "epoch": 7.308754389097185,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 30702
    },
    {
      "epoch": 7.308992441825865,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 30703
    },
    {
      "epoch": 7.309230494554543,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 30704
    },
    {
      "epoch": 7.309468547283223,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2543,
      "step": 30705
    },
    {
      "epoch": 7.309706600011903,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 30706
    },
    {
      "epoch": 7.309944652740582,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 30707
    },
    {
      "epoch": 7.310182705469262,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 30708
    },
    {
      "epoch": 7.310420758197941,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 30709
    },
    {
      "epoch": 7.31065881092662,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 30710
    },
    {
      "epoch": 7.3108968636553,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 30711
    },
    {
      "epoch": 7.311134916383979,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 30712
    },
    {
      "epoch": 7.311372969112658,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 30713
    },
    {
      "epoch": 7.311611021841338,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 30714
    },
    {
      "epoch": 7.311849074570017,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 30715
    },
    {
      "epoch": 7.312087127298697,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9186,
      "step": 30716
    },
    {
      "epoch": 7.312325180027376,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 30717
    },
    {
      "epoch": 7.312563232756055,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 30718
    },
    {
      "epoch": 7.312801285484735,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 30719
    },
    {
      "epoch": 7.3130393382134145,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 30720
    },
    {
      "epoch": 7.313277390942094,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 30721
    },
    {
      "epoch": 7.313515443670773,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 30722
    },
    {
      "epoch": 7.313753496399452,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 30723
    },
    {
      "epoch": 7.313991549128132,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0609,
      "step": 30724
    },
    {
      "epoch": 7.314229601856812,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 30725
    },
    {
      "epoch": 7.31446765458549,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 30726
    },
    {
      "epoch": 7.31470570731417,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 30727
    },
    {
      "epoch": 7.3149437600428495,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.0949,
      "step": 30728
    },
    {
      "epoch": 7.315181812771529,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 30729
    },
    {
      "epoch": 7.315419865500209,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 30730
    },
    {
      "epoch": 7.3156579182288874,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 30731
    },
    {
      "epoch": 7.315895970957567,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3789,
      "step": 30732
    },
    {
      "epoch": 7.316134023686247,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 30733
    },
    {
      "epoch": 7.316372076414926,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 1.8171,
      "step": 30734
    },
    {
      "epoch": 7.316610129143605,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 30735
    },
    {
      "epoch": 7.316848181872285,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 30736
    },
    {
      "epoch": 7.317086234600964,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 30737
    },
    {
      "epoch": 7.317324287329644,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 30738
    },
    {
      "epoch": 7.3175623400583225,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 30739
    },
    {
      "epoch": 7.317800392787002,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 30740
    },
    {
      "epoch": 7.318038445515682,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 30741
    },
    {
      "epoch": 7.318276498244361,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 30742
    },
    {
      "epoch": 7.318514550973041,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 30743
    },
    {
      "epoch": 7.31875260370172,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 30744
    },
    {
      "epoch": 7.318990656430399,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7944,
      "step": 30745
    },
    {
      "epoch": 7.319228709159079,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 30746
    },
    {
      "epoch": 7.3194667618877585,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 30747
    },
    {
      "epoch": 7.319704814616437,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 30748
    },
    {
      "epoch": 7.319942867345117,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 30749
    },
    {
      "epoch": 7.320180920073796,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 30750
    },
    {
      "epoch": 7.320418972802476,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2377,
      "step": 30751
    },
    {
      "epoch": 7.320657025531155,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 30752
    },
    {
      "epoch": 7.320895078259834,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 30753
    },
    {
      "epoch": 7.321133130988514,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 30754
    },
    {
      "epoch": 7.3213711837171935,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2459,
      "step": 30755
    },
    {
      "epoch": 7.321609236445873,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 30756
    },
    {
      "epoch": 7.321847289174552,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 30757
    },
    {
      "epoch": 7.3220853419032315,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 30758
    },
    {
      "epoch": 7.322323394631911,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 30759
    },
    {
      "epoch": 7.322561447360591,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.401,
      "step": 30760
    },
    {
      "epoch": 7.322799500089269,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 30761
    },
    {
      "epoch": 7.323037552817949,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1776,
      "step": 30762
    },
    {
      "epoch": 7.323275605546629,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 30763
    },
    {
      "epoch": 7.323513658275308,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 30764
    },
    {
      "epoch": 7.323751711003988,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4823,
      "step": 30765
    },
    {
      "epoch": 7.3239897637326665,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 30766
    },
    {
      "epoch": 7.324227816461346,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0522,
      "step": 30767
    },
    {
      "epoch": 7.324465869190026,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 30768
    },
    {
      "epoch": 7.324703921918705,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 30769
    },
    {
      "epoch": 7.324941974647384,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 30770
    },
    {
      "epoch": 7.325180027376064,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 30771
    },
    {
      "epoch": 7.325418080104743,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 30772
    },
    {
      "epoch": 7.325656132833423,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 30773
    },
    {
      "epoch": 7.325894185562102,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 30774
    },
    {
      "epoch": 7.326132238290781,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 30775
    },
    {
      "epoch": 7.326370291019461,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 30776
    },
    {
      "epoch": 7.32660834374814,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 30777
    },
    {
      "epoch": 7.32684639647682,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8476,
      "step": 30778
    },
    {
      "epoch": 7.327084449205499,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 30779
    },
    {
      "epoch": 7.327322501934178,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 30780
    },
    {
      "epoch": 7.327560554662858,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 30781
    },
    {
      "epoch": 7.3277986073915375,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 30782
    },
    {
      "epoch": 7.328036660120216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 30783
    },
    {
      "epoch": 7.328274712848896,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 30784
    },
    {
      "epoch": 7.3285127655775755,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 30785
    },
    {
      "epoch": 7.328750818306255,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 30786
    },
    {
      "epoch": 7.328988871034934,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 30787
    },
    {
      "epoch": 7.329226923763613,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 30788
    },
    {
      "epoch": 7.329464976492293,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7464,
      "step": 30789
    },
    {
      "epoch": 7.329703029220973,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 30790
    },
    {
      "epoch": 7.329941081949652,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 30791
    },
    {
      "epoch": 7.330179134678331,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 30792
    },
    {
      "epoch": 7.3304171874070105,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 30793
    },
    {
      "epoch": 7.33065524013569,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 30794
    },
    {
      "epoch": 7.33089329286437,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 30795
    },
    {
      "epoch": 7.3311313455930485,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0389,
      "step": 30796
    },
    {
      "epoch": 7.331369398321728,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 30797
    },
    {
      "epoch": 7.331607451050408,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 30798
    },
    {
      "epoch": 7.331845503779087,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7189,
      "step": 30799
    },
    {
      "epoch": 7.332083556507767,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 30800
    },
    {
      "epoch": 7.332083556507767,
      "eval_runtime": 0.5452,
      "eval_samples_per_second": 1.834,
      "eval_steps_per_second": 1.834,
      "step": 30800
    },
    {
      "epoch": 7.332321609236446,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2551,
      "step": 30801
    },
    {
      "epoch": 7.332559661965125,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 30802
    },
    {
      "epoch": 7.332797714693805,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 30803
    },
    {
      "epoch": 7.333035767422484,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 30804
    },
    {
      "epoch": 7.333273820151163,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 30805
    },
    {
      "epoch": 7.333511872879843,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 30806
    },
    {
      "epoch": 7.333749925608522,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 30807
    },
    {
      "epoch": 7.333987978337202,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 30808
    },
    {
      "epoch": 7.334226031065882,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 30809
    },
    {
      "epoch": 7.33446408379456,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 30810
    },
    {
      "epoch": 7.33470213652324,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 30811
    },
    {
      "epoch": 7.3349401892519195,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.322,
      "step": 30812
    },
    {
      "epoch": 7.335178241980599,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1911,
      "step": 30813
    },
    {
      "epoch": 7.335416294709278,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 30814
    },
    {
      "epoch": 7.335654347437957,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2192,
      "step": 30815
    },
    {
      "epoch": 7.335892400166637,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8758,
      "step": 30816
    },
    {
      "epoch": 7.336130452895317,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 30817
    },
    {
      "epoch": 7.336368505623995,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 30818
    },
    {
      "epoch": 7.336606558352675,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 30819
    },
    {
      "epoch": 7.3368446110813546,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 30820
    },
    {
      "epoch": 7.337082663810034,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 30821
    },
    {
      "epoch": 7.337320716538713,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 30822
    },
    {
      "epoch": 7.3375587692673925,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 30823
    },
    {
      "epoch": 7.337796821996072,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.738,
      "step": 30824
    },
    {
      "epoch": 7.338034874724752,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 30825
    },
    {
      "epoch": 7.338272927453431,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2229,
      "step": 30826
    },
    {
      "epoch": 7.33851098018211,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7658,
      "step": 30827
    },
    {
      "epoch": 7.33874903291079,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 30828
    },
    {
      "epoch": 7.338987085639469,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 30829
    },
    {
      "epoch": 7.339225138368149,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8427,
      "step": 30830
    },
    {
      "epoch": 7.3394631910968275,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 30831
    },
    {
      "epoch": 7.339701243825507,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 30832
    },
    {
      "epoch": 7.339939296554187,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 30833
    },
    {
      "epoch": 7.340177349282866,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 30834
    },
    {
      "epoch": 7.340415402011546,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 30835
    },
    {
      "epoch": 7.340653454740225,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3916,
      "step": 30836
    },
    {
      "epoch": 7.340891507468904,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 30837
    },
    {
      "epoch": 7.341129560197584,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 30838
    },
    {
      "epoch": 7.3413676129262635,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1693,
      "step": 30839
    },
    {
      "epoch": 7.341605665654942,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 30840
    },
    {
      "epoch": 7.341843718383622,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 30841
    },
    {
      "epoch": 7.342081771112301,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9546,
      "step": 30842
    },
    {
      "epoch": 7.342319823840981,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 30843
    },
    {
      "epoch": 7.342557876569661,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 30844
    },
    {
      "epoch": 7.342795929298339,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.8818,
      "step": 30845
    },
    {
      "epoch": 7.343033982027019,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3724,
      "step": 30846
    },
    {
      "epoch": 7.343272034755699,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 30847
    },
    {
      "epoch": 7.343510087484378,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 30848
    },
    {
      "epoch": 7.343748140213057,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 30849
    },
    {
      "epoch": 7.3439861929417365,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 30850
    },
    {
      "epoch": 7.344224245670416,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 30851
    },
    {
      "epoch": 7.344462298399096,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 30852
    },
    {
      "epoch": 7.344700351127774,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 30853
    },
    {
      "epoch": 7.344938403856454,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1635,
      "step": 30854
    },
    {
      "epoch": 7.345176456585134,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 30855
    },
    {
      "epoch": 7.345414509313813,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3153,
      "step": 30856
    },
    {
      "epoch": 7.345652562042492,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8804,
      "step": 30857
    },
    {
      "epoch": 7.345890614771172,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 30858
    },
    {
      "epoch": 7.346128667499851,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 30859
    },
    {
      "epoch": 7.346366720228531,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 30860
    },
    {
      "epoch": 7.34660477295721,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 30861
    },
    {
      "epoch": 7.346842825685889,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 30862
    },
    {
      "epoch": 7.347080878414569,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 30863
    },
    {
      "epoch": 7.347318931143248,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 30864
    },
    {
      "epoch": 7.347556983871928,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 30865
    },
    {
      "epoch": 7.347795036600607,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9183,
      "step": 30866
    },
    {
      "epoch": 7.348033089329286,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 30867
    },
    {
      "epoch": 7.348271142057966,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 30868
    },
    {
      "epoch": 7.348509194786645,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 30869
    },
    {
      "epoch": 7.348747247515325,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 30870
    },
    {
      "epoch": 7.348985300244004,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 30871
    },
    {
      "epoch": 7.349223352972683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 30872
    },
    {
      "epoch": 7.349461405701363,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 30873
    },
    {
      "epoch": 7.349699458430043,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4329,
      "step": 30874
    },
    {
      "epoch": 7.349937511158721,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 30875
    },
    {
      "epoch": 7.350175563887401,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 30876
    },
    {
      "epoch": 7.3504136166160805,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1453,
      "step": 30877
    },
    {
      "epoch": 7.35065166934476,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 30878
    },
    {
      "epoch": 7.35088972207344,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2905,
      "step": 30879
    },
    {
      "epoch": 7.351127774802118,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 30880
    },
    {
      "epoch": 7.351365827530798,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 30881
    },
    {
      "epoch": 7.351603880259478,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1586,
      "step": 30882
    },
    {
      "epoch": 7.351841932988157,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 30883
    },
    {
      "epoch": 7.352079985716836,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2694,
      "step": 30884
    },
    {
      "epoch": 7.352318038445516,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9617,
      "step": 30885
    },
    {
      "epoch": 7.352556091174195,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 30886
    },
    {
      "epoch": 7.352794143902875,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 30887
    },
    {
      "epoch": 7.3530321966315535,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 30888
    },
    {
      "epoch": 7.353270249360233,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 30889
    },
    {
      "epoch": 7.353508302088913,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7885,
      "step": 30890
    },
    {
      "epoch": 7.353746354817592,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 30891
    },
    {
      "epoch": 7.353984407546271,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 30892
    },
    {
      "epoch": 7.354222460274951,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 30893
    },
    {
      "epoch": 7.35446051300363,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2831,
      "step": 30894
    },
    {
      "epoch": 7.35469856573231,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 30895
    },
    {
      "epoch": 7.3549366184609895,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 30896
    },
    {
      "epoch": 7.355174671189668,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 30897
    },
    {
      "epoch": 7.355412723918348,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 30898
    },
    {
      "epoch": 7.355650776647027,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 30899
    },
    {
      "epoch": 7.355888829375707,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7179,
      "step": 30900
    },
    {
      "epoch": 7.355888829375707,
      "eval_runtime": 0.5289,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 30900
    },
    {
      "epoch": 7.356126882104386,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 30901
    },
    {
      "epoch": 7.356364934833065,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 30902
    },
    {
      "epoch": 7.356602987561745,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 30903
    },
    {
      "epoch": 7.3568410402904245,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.197,
      "step": 30904
    },
    {
      "epoch": 7.357079093019104,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 30905
    },
    {
      "epoch": 7.357317145747783,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 30906
    },
    {
      "epoch": 7.357555198476462,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 30907
    },
    {
      "epoch": 7.357793251205142,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 30908
    },
    {
      "epoch": 7.358031303933822,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 30909
    },
    {
      "epoch": 7.3582693566625,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 30910
    },
    {
      "epoch": 7.35850740939118,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 30911
    },
    {
      "epoch": 7.35874546211986,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 30912
    },
    {
      "epoch": 7.358983514848539,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1365,
      "step": 30913
    },
    {
      "epoch": 7.359221567577219,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 30914
    },
    {
      "epoch": 7.3594596203058975,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 30915
    },
    {
      "epoch": 7.359697673034577,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 30916
    },
    {
      "epoch": 7.359935725763257,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2787,
      "step": 30917
    },
    {
      "epoch": 7.360173778491936,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 30918
    },
    {
      "epoch": 7.360411831220615,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7898,
      "step": 30919
    },
    {
      "epoch": 7.360649883949295,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 30920
    },
    {
      "epoch": 7.360887936677974,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 30921
    },
    {
      "epoch": 7.361125989406654,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 30922
    },
    {
      "epoch": 7.361364042135333,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 30923
    },
    {
      "epoch": 7.361602094864012,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 30924
    },
    {
      "epoch": 7.361840147592692,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9001,
      "step": 30925
    },
    {
      "epoch": 7.362078200321371,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 30926
    },
    {
      "epoch": 7.362316253050051,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 30927
    },
    {
      "epoch": 7.36255430577873,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 30928
    },
    {
      "epoch": 7.362792358507409,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 30929
    },
    {
      "epoch": 7.363030411236089,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 30930
    },
    {
      "epoch": 7.3632684639647685,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 30931
    },
    {
      "epoch": 7.363506516693447,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 30932
    },
    {
      "epoch": 7.363744569422127,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1739,
      "step": 30933
    },
    {
      "epoch": 7.3639826221508065,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.7551,
      "step": 30934
    },
    {
      "epoch": 7.364220674879486,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 30935
    },
    {
      "epoch": 7.364458727608165,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2915,
      "step": 30936
    },
    {
      "epoch": 7.364696780336844,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 30937
    },
    {
      "epoch": 7.364934833065524,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 30938
    },
    {
      "epoch": 7.365172885794204,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.283,
      "step": 30939
    },
    {
      "epoch": 7.365410938522883,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 30940
    },
    {
      "epoch": 7.365648991251562,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 30941
    },
    {
      "epoch": 7.3658870439802415,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 30942
    },
    {
      "epoch": 7.366125096708921,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1123,
      "step": 30943
    },
    {
      "epoch": 7.366363149437601,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 30944
    },
    {
      "epoch": 7.3666012021662794,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 30945
    },
    {
      "epoch": 7.366839254894959,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7837,
      "step": 30946
    },
    {
      "epoch": 7.367077307623639,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 30947
    },
    {
      "epoch": 7.367315360352318,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 30948
    },
    {
      "epoch": 7.367553413080998,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 30949
    },
    {
      "epoch": 7.367791465809677,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6188,
      "step": 30950
    },
    {
      "epoch": 7.368029518538356,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 30951
    },
    {
      "epoch": 7.368267571267036,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.824,
      "step": 30952
    },
    {
      "epoch": 7.368505623995715,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 30953
    },
    {
      "epoch": 7.368743676724394,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 30954
    },
    {
      "epoch": 7.368981729453074,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.208,
      "step": 30955
    },
    {
      "epoch": 7.369219782181753,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 30956
    },
    {
      "epoch": 7.369457834910433,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 30957
    },
    {
      "epoch": 7.369695887639112,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 30958
    },
    {
      "epoch": 7.369933940367791,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 30959
    },
    {
      "epoch": 7.370171993096471,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 30960
    },
    {
      "epoch": 7.3704100458251505,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.4891,
      "step": 30961
    },
    {
      "epoch": 7.37064809855383,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 30962
    },
    {
      "epoch": 7.370886151282509,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 30963
    },
    {
      "epoch": 7.371124204011188,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 30964
    },
    {
      "epoch": 7.371362256739868,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2662,
      "step": 30965
    },
    {
      "epoch": 7.371600309468548,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9821,
      "step": 30966
    },
    {
      "epoch": 7.371838362197226,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.644,
      "step": 30967
    },
    {
      "epoch": 7.372076414925906,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3256,
      "step": 30968
    },
    {
      "epoch": 7.3723144676545855,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 30969
    },
    {
      "epoch": 7.372552520383265,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 30970
    },
    {
      "epoch": 7.372790573111944,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 30971
    },
    {
      "epoch": 7.3730286258406235,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 30972
    },
    {
      "epoch": 7.373266678569303,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 30973
    },
    {
      "epoch": 7.373504731297983,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 30974
    },
    {
      "epoch": 7.373742784026662,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 30975
    },
    {
      "epoch": 7.373980836755341,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 30976
    },
    {
      "epoch": 7.374218889484021,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0206,
      "step": 30977
    },
    {
      "epoch": 7.3744569422127,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 30978
    },
    {
      "epoch": 7.37469499494138,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 30979
    },
    {
      "epoch": 7.3749330476700585,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2813,
      "step": 30980
    },
    {
      "epoch": 7.375171100398738,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 30981
    },
    {
      "epoch": 7.375409153127418,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 30982
    },
    {
      "epoch": 7.375647205856097,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7676,
      "step": 30983
    },
    {
      "epoch": 7.375885258584777,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 30984
    },
    {
      "epoch": 7.376123311313456,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 30985
    },
    {
      "epoch": 7.376361364042135,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9287,
      "step": 30986
    },
    {
      "epoch": 7.376599416770815,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 30987
    },
    {
      "epoch": 7.3768374694994945,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 30988
    },
    {
      "epoch": 7.377075522228173,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3878,
      "step": 30989
    },
    {
      "epoch": 7.377313574956853,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 30990
    },
    {
      "epoch": 7.377551627685532,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 30991
    },
    {
      "epoch": 7.377789680414212,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 30992
    },
    {
      "epoch": 7.378027733142892,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.999,
      "step": 30993
    },
    {
      "epoch": 7.37826578587157,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0243,
      "step": 30994
    },
    {
      "epoch": 7.37850383860025,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 30995
    },
    {
      "epoch": 7.3787418913289295,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0895,
      "step": 30996
    },
    {
      "epoch": 7.378979944057609,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.527,
      "step": 30997
    },
    {
      "epoch": 7.379217996786288,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 30998
    },
    {
      "epoch": 7.3794560495149675,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2932,
      "step": 30999
    },
    {
      "epoch": 7.379694102243647,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 31000
    },
    {
      "epoch": 7.379694102243647,
      "eval_runtime": 0.5313,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 31000
    },
    {
      "epoch": 7.379932154972327,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7369,
      "step": 31001
    },
    {
      "epoch": 7.380170207701005,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 31002
    },
    {
      "epoch": 7.380408260429685,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 31003
    },
    {
      "epoch": 7.380646313158365,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 31004
    },
    {
      "epoch": 7.380884365887044,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 31005
    },
    {
      "epoch": 7.381122418615723,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 31006
    },
    {
      "epoch": 7.3813604713444025,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 31007
    },
    {
      "epoch": 7.381598524073082,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 31008
    },
    {
      "epoch": 7.381836576801762,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 31009
    },
    {
      "epoch": 7.382074629530441,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 31010
    },
    {
      "epoch": 7.38231268225912,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 31011
    },
    {
      "epoch": 7.3825507349878,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0013,
      "step": 31012
    },
    {
      "epoch": 7.382788787716479,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 31013
    },
    {
      "epoch": 7.383026840445159,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3217,
      "step": 31014
    },
    {
      "epoch": 7.383264893173838,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 31015
    },
    {
      "epoch": 7.383502945902517,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 31016
    },
    {
      "epoch": 7.383740998631197,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 31017
    },
    {
      "epoch": 7.383979051359876,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 31018
    },
    {
      "epoch": 7.384217104088556,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 31019
    },
    {
      "epoch": 7.384455156817235,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 31020
    },
    {
      "epoch": 7.384693209545914,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8739,
      "step": 31021
    },
    {
      "epoch": 7.384931262274594,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 31022
    },
    {
      "epoch": 7.385169315003274,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9773,
      "step": 31023
    },
    {
      "epoch": 7.385407367731952,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 31024
    },
    {
      "epoch": 7.385645420460632,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 31025
    },
    {
      "epoch": 7.3858834731893115,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 31026
    },
    {
      "epoch": 7.386121525917991,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 31027
    },
    {
      "epoch": 7.386359578646671,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 31028
    },
    {
      "epoch": 7.386597631375349,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 31029
    },
    {
      "epoch": 7.386835684104029,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 31030
    },
    {
      "epoch": 7.387073736832709,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 31031
    },
    {
      "epoch": 7.387311789561388,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 31032
    },
    {
      "epoch": 7.387549842290067,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 31033
    },
    {
      "epoch": 7.3877878950187466,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3422,
      "step": 31034
    },
    {
      "epoch": 7.388025947747426,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 31035
    },
    {
      "epoch": 7.388264000476106,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 31036
    },
    {
      "epoch": 7.3885020532047845,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 31037
    },
    {
      "epoch": 7.388740105933464,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.3001,
      "step": 31038
    },
    {
      "epoch": 7.388978158662144,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 31039
    },
    {
      "epoch": 7.389216211390823,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0439,
      "step": 31040
    },
    {
      "epoch": 7.389454264119502,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2928,
      "step": 31041
    },
    {
      "epoch": 7.389692316848182,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 31042
    },
    {
      "epoch": 7.389930369576861,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 31043
    },
    {
      "epoch": 7.390168422305541,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 31044
    },
    {
      "epoch": 7.39040647503422,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 31045
    },
    {
      "epoch": 7.390644527762899,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 31046
    },
    {
      "epoch": 7.390882580491579,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 31047
    },
    {
      "epoch": 7.391120633220258,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 31048
    },
    {
      "epoch": 7.391358685948938,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 31049
    },
    {
      "epoch": 7.391596738677617,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 31050
    },
    {
      "epoch": 7.391834791406296,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 31051
    },
    {
      "epoch": 7.392072844134976,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 31052
    },
    {
      "epoch": 7.3923108968636555,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.308,
      "step": 31053
    },
    {
      "epoch": 7.392548949592335,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 31054
    },
    {
      "epoch": 7.392787002321014,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 31055
    },
    {
      "epoch": 7.393025055049693,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 31056
    },
    {
      "epoch": 7.393263107778373,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.169,
      "step": 31057
    },
    {
      "epoch": 7.393501160507053,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 31058
    },
    {
      "epoch": 7.393739213235731,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 31059
    },
    {
      "epoch": 7.393977265964411,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 31060
    },
    {
      "epoch": 7.394215318693091,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 31061
    },
    {
      "epoch": 7.39445337142177,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 31062
    },
    {
      "epoch": 7.39469142415045,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 31063
    },
    {
      "epoch": 7.3949294768791285,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 31064
    },
    {
      "epoch": 7.395167529607808,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 31065
    },
    {
      "epoch": 7.395405582336488,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 31066
    },
    {
      "epoch": 7.395643635065167,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 31067
    },
    {
      "epoch": 7.395881687793846,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 31068
    },
    {
      "epoch": 7.396119740522526,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 31069
    },
    {
      "epoch": 7.396357793251205,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 31070
    },
    {
      "epoch": 7.396595845979885,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 31071
    },
    {
      "epoch": 7.396833898708564,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.6504,
      "step": 31072
    },
    {
      "epoch": 7.397071951437243,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 31073
    },
    {
      "epoch": 7.397310004165923,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 31074
    },
    {
      "epoch": 7.397548056894602,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 31075
    },
    {
      "epoch": 7.397786109623281,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 31076
    },
    {
      "epoch": 7.398024162351961,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 31077
    },
    {
      "epoch": 7.39826221508064,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 31078
    },
    {
      "epoch": 7.39850026780932,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 31079
    },
    {
      "epoch": 7.3987383205379995,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 31080
    },
    {
      "epoch": 7.398976373266678,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 31081
    },
    {
      "epoch": 7.399214425995358,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 31082
    },
    {
      "epoch": 7.399452478724037,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8272,
      "step": 31083
    },
    {
      "epoch": 7.399690531452717,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 31084
    },
    {
      "epoch": 7.399928584181396,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 31085
    },
    {
      "epoch": 7.400166636910075,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 31086
    },
    {
      "epoch": 7.400404689638755,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 31087
    },
    {
      "epoch": 7.400642742367435,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 31088
    },
    {
      "epoch": 7.400880795096114,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 31089
    },
    {
      "epoch": 7.401118847824793,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 31090
    },
    {
      "epoch": 7.4013569005534725,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 31091
    },
    {
      "epoch": 7.401594953282152,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 31092
    },
    {
      "epoch": 7.401833006010832,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 31093
    },
    {
      "epoch": 7.40207105873951,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 31094
    },
    {
      "epoch": 7.40230911146819,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 31095
    },
    {
      "epoch": 7.40254716419687,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6542,
      "step": 31096
    },
    {
      "epoch": 7.402785216925549,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 31097
    },
    {
      "epoch": 7.403023269654229,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.6531,
      "step": 31098
    },
    {
      "epoch": 7.403261322382908,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 31099
    },
    {
      "epoch": 7.403499375111587,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 31100
    },
    {
      "epoch": 7.403499375111587,
      "eval_runtime": 0.5211,
      "eval_samples_per_second": 1.919,
      "eval_steps_per_second": 1.919,
      "step": 31100
    },
    {
      "epoch": 7.403737427840267,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.45,
      "step": 31101
    },
    {
      "epoch": 7.403975480568946,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 31102
    },
    {
      "epoch": 7.404213533297625,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.4528,
      "step": 31103
    },
    {
      "epoch": 7.404451586026305,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 31104
    },
    {
      "epoch": 7.404689638754984,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 31105
    },
    {
      "epoch": 7.404927691483664,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 31106
    },
    {
      "epoch": 7.405165744212343,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 31107
    },
    {
      "epoch": 7.405403796941022,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 31108
    },
    {
      "epoch": 7.405641849669702,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 31109
    },
    {
      "epoch": 7.4058799023983815,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.4237,
      "step": 31110
    },
    {
      "epoch": 7.406117955127061,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.171,
      "step": 31111
    },
    {
      "epoch": 7.40635600785574,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 31112
    },
    {
      "epoch": 7.406594060584419,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 31113
    },
    {
      "epoch": 7.406832113313099,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 31114
    },
    {
      "epoch": 7.407070166041779,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 31115
    },
    {
      "epoch": 7.407308218770457,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 31116
    },
    {
      "epoch": 7.407546271499137,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 31117
    },
    {
      "epoch": 7.4077843242278165,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0043,
      "step": 31118
    },
    {
      "epoch": 7.408022376956496,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2132,
      "step": 31119
    },
    {
      "epoch": 7.408260429685175,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 31120
    },
    {
      "epoch": 7.4084984824138544,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 31121
    },
    {
      "epoch": 7.408736535142534,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 31122
    },
    {
      "epoch": 7.408974587871214,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 31123
    },
    {
      "epoch": 7.409212640599893,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 31124
    },
    {
      "epoch": 7.409450693328572,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2642,
      "step": 31125
    },
    {
      "epoch": 7.409688746057252,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 31126
    },
    {
      "epoch": 7.409926798785931,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 31127
    },
    {
      "epoch": 7.410164851514611,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2463,
      "step": 31128
    },
    {
      "epoch": 7.4104029042432895,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 31129
    },
    {
      "epoch": 7.410640956971969,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 31130
    },
    {
      "epoch": 7.410879009700649,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 2.1881,
      "step": 31131
    },
    {
      "epoch": 7.411117062429328,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 31132
    },
    {
      "epoch": 7.411355115158008,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 31133
    },
    {
      "epoch": 7.411593167886687,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 31134
    },
    {
      "epoch": 7.411831220615366,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 31135
    },
    {
      "epoch": 7.412069273344046,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.966,
      "step": 31136
    },
    {
      "epoch": 7.4123073260727255,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 31137
    },
    {
      "epoch": 7.412545378801404,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.98,
      "step": 31138
    },
    {
      "epoch": 7.412783431530084,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 31139
    },
    {
      "epoch": 7.413021484258763,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 31140
    },
    {
      "epoch": 7.413259536987443,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2261,
      "step": 31141
    },
    {
      "epoch": 7.413497589716122,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 31142
    },
    {
      "epoch": 7.413735642444801,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 31143
    },
    {
      "epoch": 7.413973695173481,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 31144
    },
    {
      "epoch": 7.4142117479021605,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 31145
    },
    {
      "epoch": 7.41444980063084,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 31146
    },
    {
      "epoch": 7.414687853359519,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 31147
    },
    {
      "epoch": 7.4149259060881985,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 31148
    },
    {
      "epoch": 7.415163958816878,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1401,
      "step": 31149
    },
    {
      "epoch": 7.415402011545558,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 31150
    },
    {
      "epoch": 7.415640064274236,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7895,
      "step": 31151
    },
    {
      "epoch": 7.415878117002916,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3218,
      "step": 31152
    },
    {
      "epoch": 7.416116169731596,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 31153
    },
    {
      "epoch": 7.416354222460275,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 31154
    },
    {
      "epoch": 7.416592275188954,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 31155
    },
    {
      "epoch": 7.4168303279176335,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 31156
    },
    {
      "epoch": 7.417068380646313,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 31157
    },
    {
      "epoch": 7.417306433374993,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 31158
    },
    {
      "epoch": 7.417544486103672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 31159
    },
    {
      "epoch": 7.417782538832351,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 31160
    },
    {
      "epoch": 7.418020591561031,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 31161
    },
    {
      "epoch": 7.41825864428971,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 31162
    },
    {
      "epoch": 7.41849669701839,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 31163
    },
    {
      "epoch": 7.418734749747069,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 31164
    },
    {
      "epoch": 7.418972802475748,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 31165
    },
    {
      "epoch": 7.419210855204428,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 31166
    },
    {
      "epoch": 7.419448907933107,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 31167
    },
    {
      "epoch": 7.419686960661787,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 31168
    },
    {
      "epoch": 7.419925013390466,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2133,
      "step": 31169
    },
    {
      "epoch": 7.420163066119145,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 31170
    },
    {
      "epoch": 7.420401118847825,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 31171
    },
    {
      "epoch": 7.4206391715765045,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4375,
      "step": 31172
    },
    {
      "epoch": 7.420877224305183,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8499,
      "step": 31173
    },
    {
      "epoch": 7.421115277033863,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 31174
    },
    {
      "epoch": 7.4213533297625425,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 31175
    },
    {
      "epoch": 7.421591382491222,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2964,
      "step": 31176
    },
    {
      "epoch": 7.421829435219901,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0661,
      "step": 31177
    },
    {
      "epoch": 7.42206748794858,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 1.801,
      "step": 31178
    },
    {
      "epoch": 7.42230554067726,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 31179
    },
    {
      "epoch": 7.42254359340594,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 31180
    },
    {
      "epoch": 7.422781646134619,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 31181
    },
    {
      "epoch": 7.423019698863298,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7749,
      "step": 31182
    },
    {
      "epoch": 7.4232577515919775,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 31183
    },
    {
      "epoch": 7.423495804320657,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0588,
      "step": 31184
    },
    {
      "epoch": 7.423733857049337,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7334,
      "step": 31185
    },
    {
      "epoch": 7.4239719097780155,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 31186
    },
    {
      "epoch": 7.424209962506695,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 31187
    },
    {
      "epoch": 7.424448015235375,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 31188
    },
    {
      "epoch": 7.424686067964054,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 31189
    },
    {
      "epoch": 7.424924120692733,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 31190
    },
    {
      "epoch": 7.425162173421413,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3744,
      "step": 31191
    },
    {
      "epoch": 7.425400226150092,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 31192
    },
    {
      "epoch": 7.425638278878772,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 31193
    },
    {
      "epoch": 7.425876331607451,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 31194
    },
    {
      "epoch": 7.42611438433613,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 31195
    },
    {
      "epoch": 7.42635243706481,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 31196
    },
    {
      "epoch": 7.426590489793489,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 31197
    },
    {
      "epoch": 7.426828542522169,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 31198
    },
    {
      "epoch": 7.427066595250848,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9058,
      "step": 31199
    },
    {
      "epoch": 7.427304647979527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 31200
    },
    {
      "epoch": 7.427304647979527,
      "eval_runtime": 0.5275,
      "eval_samples_per_second": 1.896,
      "eval_steps_per_second": 1.896,
      "step": 31200
    },
    {
      "epoch": 7.427542700708207,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 31201
    },
    {
      "epoch": 7.4277807534368865,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 31202
    },
    {
      "epoch": 7.428018806165566,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 31203
    },
    {
      "epoch": 7.428256858894245,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 31204
    },
    {
      "epoch": 7.428494911622924,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 31205
    },
    {
      "epoch": 7.428732964351604,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 31206
    },
    {
      "epoch": 7.428971017080284,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.85,
      "step": 31207
    },
    {
      "epoch": 7.429209069808962,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2863,
      "step": 31208
    },
    {
      "epoch": 7.429447122537642,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 31209
    },
    {
      "epoch": 7.4296851752663216,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 31210
    },
    {
      "epoch": 7.429923227995001,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 31211
    },
    {
      "epoch": 7.430161280723681,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 31212
    },
    {
      "epoch": 7.4303993334523595,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 31213
    },
    {
      "epoch": 7.430637386181039,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 31214
    },
    {
      "epoch": 7.430875438909719,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 31215
    },
    {
      "epoch": 7.431113491638398,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 31216
    },
    {
      "epoch": 7.431351544367077,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4077,
      "step": 31217
    },
    {
      "epoch": 7.431589597095757,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 31218
    },
    {
      "epoch": 7.431827649824436,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 31219
    },
    {
      "epoch": 7.432065702553116,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 31220
    },
    {
      "epoch": 7.4323037552817945,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 31221
    },
    {
      "epoch": 7.432541808010474,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 31222
    },
    {
      "epoch": 7.432779860739154,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 31223
    },
    {
      "epoch": 7.433017913467833,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 31224
    },
    {
      "epoch": 7.433255966196512,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 31225
    },
    {
      "epoch": 7.433494018925192,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 31226
    },
    {
      "epoch": 7.433732071653871,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 31227
    },
    {
      "epoch": 7.433970124382551,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8974,
      "step": 31228
    },
    {
      "epoch": 7.4342081771112305,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 31229
    },
    {
      "epoch": 7.434446229839909,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 31230
    },
    {
      "epoch": 7.434684282568589,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 31231
    },
    {
      "epoch": 7.434922335297268,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7623,
      "step": 31232
    },
    {
      "epoch": 7.435160388025948,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 31233
    },
    {
      "epoch": 7.435398440754627,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 31234
    },
    {
      "epoch": 7.435636493483306,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 31235
    },
    {
      "epoch": 7.435874546211986,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2096,
      "step": 31236
    },
    {
      "epoch": 7.436112598940666,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 31237
    },
    {
      "epoch": 7.436350651669345,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 31238
    },
    {
      "epoch": 7.436588704398024,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2138,
      "step": 31239
    },
    {
      "epoch": 7.4368267571267035,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 31240
    },
    {
      "epoch": 7.437064809855383,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 31241
    },
    {
      "epoch": 7.437302862584063,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.6271,
      "step": 31242
    },
    {
      "epoch": 7.437540915312741,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 31243
    },
    {
      "epoch": 7.437778968041421,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 31244
    },
    {
      "epoch": 7.438017020770101,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 31245
    },
    {
      "epoch": 7.43825507349878,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 31246
    },
    {
      "epoch": 7.43849312622746,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 31247
    },
    {
      "epoch": 7.438731178956139,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9116,
      "step": 31248
    },
    {
      "epoch": 7.438969231684818,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3689,
      "step": 31249
    },
    {
      "epoch": 7.439207284413498,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1983,
      "step": 31250
    },
    {
      "epoch": 7.439445337142177,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 31251
    },
    {
      "epoch": 7.439683389870856,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 31252
    },
    {
      "epoch": 7.439921442599536,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3039,
      "step": 31253
    },
    {
      "epoch": 7.440159495328215,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 31254
    },
    {
      "epoch": 7.440397548056895,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 31255
    },
    {
      "epoch": 7.440635600785574,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0504,
      "step": 31256
    },
    {
      "epoch": 7.440873653514253,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 31257
    },
    {
      "epoch": 7.441111706242933,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 31258
    },
    {
      "epoch": 7.441349758971612,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 31259
    },
    {
      "epoch": 7.441587811700291,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3491,
      "step": 31260
    },
    {
      "epoch": 7.441825864428971,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 31261
    },
    {
      "epoch": 7.44206391715765,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 31262
    },
    {
      "epoch": 7.44230196988633,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 31263
    },
    {
      "epoch": 7.44254002261501,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 31264
    },
    {
      "epoch": 7.442778075343688,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 31265
    },
    {
      "epoch": 7.443016128072368,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 31266
    },
    {
      "epoch": 7.4432541808010475,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2711,
      "step": 31267
    },
    {
      "epoch": 7.443492233529727,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.988,
      "step": 31268
    },
    {
      "epoch": 7.443730286258406,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 31269
    },
    {
      "epoch": 7.443968338987085,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 31270
    },
    {
      "epoch": 7.444206391715765,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 31271
    },
    {
      "epoch": 7.444444444444445,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2052,
      "step": 31272
    },
    {
      "epoch": 7.444682497173124,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 31273
    },
    {
      "epoch": 7.444920549901803,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 31274
    },
    {
      "epoch": 7.445158602630483,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 31275
    },
    {
      "epoch": 7.445396655359162,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 31276
    },
    {
      "epoch": 7.445634708087842,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 31277
    },
    {
      "epoch": 7.4458727608165205,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3956,
      "step": 31278
    },
    {
      "epoch": 7.4461108135452,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 31279
    },
    {
      "epoch": 7.44634886627388,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 31280
    },
    {
      "epoch": 7.446586919002559,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0507,
      "step": 31281
    },
    {
      "epoch": 7.446824971731239,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 31282
    },
    {
      "epoch": 7.447063024459918,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 31283
    },
    {
      "epoch": 7.447301077188597,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 31284
    },
    {
      "epoch": 7.447539129917277,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8189,
      "step": 31285
    },
    {
      "epoch": 7.4477771826459565,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 31286
    },
    {
      "epoch": 7.448015235374635,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 31287
    },
    {
      "epoch": 7.448253288103315,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 31288
    },
    {
      "epoch": 7.448491340831994,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 31289
    },
    {
      "epoch": 7.448729393560674,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 31290
    },
    {
      "epoch": 7.448967446289353,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7661,
      "step": 31291
    },
    {
      "epoch": 7.449205499018032,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 31292
    },
    {
      "epoch": 7.449443551746712,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 31293
    },
    {
      "epoch": 7.4496816044753915,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 31294
    },
    {
      "epoch": 7.44991965720407,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 31295
    },
    {
      "epoch": 7.45015770993275,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1266,
      "step": 31296
    },
    {
      "epoch": 7.450395762661429,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 31297
    },
    {
      "epoch": 7.450633815390109,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 31298
    },
    {
      "epoch": 7.450871868118789,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 31299
    },
    {
      "epoch": 7.451109920847467,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 31300
    },
    {
      "epoch": 7.451109920847467,
      "eval_runtime": 0.5234,
      "eval_samples_per_second": 1.911,
      "eval_steps_per_second": 1.911,
      "step": 31300
    },
    {
      "epoch": 7.451347973576147,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 31301
    },
    {
      "epoch": 7.451586026304827,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 31302
    },
    {
      "epoch": 7.451824079033506,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 31303
    },
    {
      "epoch": 7.452062131762185,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 31304
    },
    {
      "epoch": 7.4523001844908645,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 31305
    },
    {
      "epoch": 7.452538237219544,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 31306
    },
    {
      "epoch": 7.452776289948224,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 31307
    },
    {
      "epoch": 7.453014342676903,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8971,
      "step": 31308
    },
    {
      "epoch": 7.453252395405582,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9657,
      "step": 31309
    },
    {
      "epoch": 7.453490448134262,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 31310
    },
    {
      "epoch": 7.453728500862941,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 31311
    },
    {
      "epoch": 7.453966553591621,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 31312
    },
    {
      "epoch": 7.4542046063203,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2437,
      "step": 31313
    },
    {
      "epoch": 7.454442659048979,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 31314
    },
    {
      "epoch": 7.454680711777659,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0992,
      "step": 31315
    },
    {
      "epoch": 7.454918764506338,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 31316
    },
    {
      "epoch": 7.455156817235018,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 31317
    },
    {
      "epoch": 7.455394869963697,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 31318
    },
    {
      "epoch": 7.455632922692376,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 31319
    },
    {
      "epoch": 7.455870975421056,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.145,
      "step": 31320
    },
    {
      "epoch": 7.4561090281497355,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 31321
    },
    {
      "epoch": 7.456347080878414,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 31322
    },
    {
      "epoch": 7.456585133607094,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 31323
    },
    {
      "epoch": 7.4568231863357735,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 31324
    },
    {
      "epoch": 7.457061239064453,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0607,
      "step": 31325
    },
    {
      "epoch": 7.457299291793132,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 31326
    },
    {
      "epoch": 7.457537344521811,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 31327
    },
    {
      "epoch": 7.457775397250491,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 31328
    },
    {
      "epoch": 7.458013449979171,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 31329
    },
    {
      "epoch": 7.45825150270785,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 31330
    },
    {
      "epoch": 7.458489555436529,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 31331
    },
    {
      "epoch": 7.4587276081652085,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8728,
      "step": 31332
    },
    {
      "epoch": 7.458965660893888,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 31333
    },
    {
      "epoch": 7.459203713622568,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.517,
      "step": 31334
    },
    {
      "epoch": 7.4594417663512465,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 31335
    },
    {
      "epoch": 7.459679819079926,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 31336
    },
    {
      "epoch": 7.459917871808606,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 31337
    },
    {
      "epoch": 7.460155924537285,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 31338
    },
    {
      "epoch": 7.460393977265964,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 31339
    },
    {
      "epoch": 7.460632029994644,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.6219,
      "step": 31340
    },
    {
      "epoch": 7.460870082723323,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2029,
      "step": 31341
    },
    {
      "epoch": 7.461108135452003,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 31342
    },
    {
      "epoch": 7.461346188180682,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 31343
    },
    {
      "epoch": 7.461584240909361,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 31344
    },
    {
      "epoch": 7.461822293638041,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 31345
    },
    {
      "epoch": 7.46206034636672,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 31346
    },
    {
      "epoch": 7.4622983990954,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 31347
    },
    {
      "epoch": 7.462536451824079,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 31348
    },
    {
      "epoch": 7.462774504552758,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0601,
      "step": 31349
    },
    {
      "epoch": 7.463012557281438,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3002,
      "step": 31350
    },
    {
      "epoch": 7.4632506100101175,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 31351
    },
    {
      "epoch": 7.463488662738797,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 31352
    },
    {
      "epoch": 7.463726715467476,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 31353
    },
    {
      "epoch": 7.463964768196155,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 31354
    },
    {
      "epoch": 7.464202820924835,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 31355
    },
    {
      "epoch": 7.464440873653515,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 31356
    },
    {
      "epoch": 7.464678926382193,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 31357
    },
    {
      "epoch": 7.464916979110873,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 31358
    },
    {
      "epoch": 7.4651550318395525,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 31359
    },
    {
      "epoch": 7.465393084568232,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1802,
      "step": 31360
    },
    {
      "epoch": 7.465631137296911,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7138,
      "step": 31361
    },
    {
      "epoch": 7.4658691900255905,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 31362
    },
    {
      "epoch": 7.46610724275427,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 31363
    },
    {
      "epoch": 7.46634529548295,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9505,
      "step": 31364
    },
    {
      "epoch": 7.466583348211629,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1767,
      "step": 31365
    },
    {
      "epoch": 7.466821400940308,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.7098,
      "step": 31366
    },
    {
      "epoch": 7.467059453668988,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 31367
    },
    {
      "epoch": 7.467297506397667,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.337,
      "step": 31368
    },
    {
      "epoch": 7.467535559126347,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 31369
    },
    {
      "epoch": 7.4677736118550255,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 31370
    },
    {
      "epoch": 7.468011664583705,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9266,
      "step": 31371
    },
    {
      "epoch": 7.468249717312385,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7869,
      "step": 31372
    },
    {
      "epoch": 7.468487770041064,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 31373
    },
    {
      "epoch": 7.468725822769743,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 31374
    },
    {
      "epoch": 7.468963875498423,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 31375
    },
    {
      "epoch": 7.469201928227102,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 31376
    },
    {
      "epoch": 7.469439980955782,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9874,
      "step": 31377
    },
    {
      "epoch": 7.4696780336844615,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 31378
    },
    {
      "epoch": 7.46991608641314,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 31379
    },
    {
      "epoch": 7.47015413914182,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 31380
    },
    {
      "epoch": 7.470392191870499,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 31381
    },
    {
      "epoch": 7.470630244599179,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3601,
      "step": 31382
    },
    {
      "epoch": 7.470868297327858,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 31383
    },
    {
      "epoch": 7.471106350056537,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 31384
    },
    {
      "epoch": 7.471344402785217,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2289,
      "step": 31385
    },
    {
      "epoch": 7.4715824555138965,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 31386
    },
    {
      "epoch": 7.471820508242576,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9475,
      "step": 31387
    },
    {
      "epoch": 7.472058560971255,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 31388
    },
    {
      "epoch": 7.4722966136999345,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 31389
    },
    {
      "epoch": 7.472534666428614,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 31390
    },
    {
      "epoch": 7.472772719157294,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.6985,
      "step": 31391
    },
    {
      "epoch": 7.473010771885972,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3348,
      "step": 31392
    },
    {
      "epoch": 7.473248824614652,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 31393
    },
    {
      "epoch": 7.473486877343332,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 31394
    },
    {
      "epoch": 7.473724930072011,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 31395
    },
    {
      "epoch": 7.473962982800691,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 31396
    },
    {
      "epoch": 7.4742010355293695,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 31397
    },
    {
      "epoch": 7.474439088258049,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.5449,
      "step": 31398
    },
    {
      "epoch": 7.474677140986729,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2141,
      "step": 31399
    },
    {
      "epoch": 7.474915193715408,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 31400
    },
    {
      "epoch": 7.474915193715408,
      "eval_runtime": 0.5303,
      "eval_samples_per_second": 1.886,
      "eval_steps_per_second": 1.886,
      "step": 31400
    },
    {
      "epoch": 7.475153246444087,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8642,
      "step": 31401
    },
    {
      "epoch": 7.475391299172767,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1172,
      "step": 31402
    },
    {
      "epoch": 7.475629351901446,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9427,
      "step": 31403
    },
    {
      "epoch": 7.475867404630126,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 31404
    },
    {
      "epoch": 7.476105457358805,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 31405
    },
    {
      "epoch": 7.476343510087484,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 31406
    },
    {
      "epoch": 7.476581562816164,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 31407
    },
    {
      "epoch": 7.476819615544843,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1574,
      "step": 31408
    },
    {
      "epoch": 7.477057668273522,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0674,
      "step": 31409
    },
    {
      "epoch": 7.477295721002202,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 31410
    },
    {
      "epoch": 7.477533773730881,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0649,
      "step": 31411
    },
    {
      "epoch": 7.477771826459561,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 31412
    },
    {
      "epoch": 7.478009879188241,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 31413
    },
    {
      "epoch": 7.478247931916919,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 31414
    },
    {
      "epoch": 7.478485984645599,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 31415
    },
    {
      "epoch": 7.4787240373742785,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2466,
      "step": 31416
    },
    {
      "epoch": 7.478962090102958,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 31417
    },
    {
      "epoch": 7.479200142831637,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 31418
    },
    {
      "epoch": 7.479438195560316,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 31419
    },
    {
      "epoch": 7.479676248288996,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 31420
    },
    {
      "epoch": 7.479914301017676,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 31421
    },
    {
      "epoch": 7.480152353746355,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 31422
    },
    {
      "epoch": 7.480390406475034,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 31423
    },
    {
      "epoch": 7.4806284592037136,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 31424
    },
    {
      "epoch": 7.480866511932393,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4533,
      "step": 31425
    },
    {
      "epoch": 7.481104564661073,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 31426
    },
    {
      "epoch": 7.4813426173897515,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.2755,
      "step": 31427
    },
    {
      "epoch": 7.481580670118431,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 31428
    },
    {
      "epoch": 7.481818722847111,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 31429
    },
    {
      "epoch": 7.48205677557579,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8154,
      "step": 31430
    },
    {
      "epoch": 7.48229482830447,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 31431
    },
    {
      "epoch": 7.482532881033149,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 31432
    },
    {
      "epoch": 7.482770933761828,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 31433
    },
    {
      "epoch": 7.483008986490508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 31434
    },
    {
      "epoch": 7.483247039219187,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 31435
    },
    {
      "epoch": 7.483485091947866,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 31436
    },
    {
      "epoch": 7.483723144676546,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 31437
    },
    {
      "epoch": 7.483961197405225,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 31438
    },
    {
      "epoch": 7.484199250133905,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 31439
    },
    {
      "epoch": 7.484437302862584,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 31440
    },
    {
      "epoch": 7.484675355591263,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.307,
      "step": 31441
    },
    {
      "epoch": 7.484913408319943,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 31442
    },
    {
      "epoch": 7.4851514610486225,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 31443
    },
    {
      "epoch": 7.485389513777301,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 31444
    },
    {
      "epoch": 7.485627566505981,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 31445
    },
    {
      "epoch": 7.48586561923466,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 31446
    },
    {
      "epoch": 7.48610367196334,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.269,
      "step": 31447
    },
    {
      "epoch": 7.48634172469202,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.969,
      "step": 31448
    },
    {
      "epoch": 7.486579777420698,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 31449
    },
    {
      "epoch": 7.486817830149378,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 31450
    },
    {
      "epoch": 7.487055882878058,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1947,
      "step": 31451
    },
    {
      "epoch": 7.487293935606737,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2447,
      "step": 31452
    },
    {
      "epoch": 7.487531988335416,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 31453
    },
    {
      "epoch": 7.4877700410640955,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 31454
    },
    {
      "epoch": 7.488008093792775,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 31455
    },
    {
      "epoch": 7.488246146521455,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 31456
    },
    {
      "epoch": 7.488484199250134,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 31457
    },
    {
      "epoch": 7.488722251978813,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 31458
    },
    {
      "epoch": 7.488960304707493,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2398,
      "step": 31459
    },
    {
      "epoch": 7.489198357436172,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1957,
      "step": 31460
    },
    {
      "epoch": 7.489436410164852,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2127,
      "step": 31461
    },
    {
      "epoch": 7.489674462893531,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 31462
    },
    {
      "epoch": 7.48991251562221,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 31463
    },
    {
      "epoch": 7.49015056835089,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4785,
      "step": 31464
    },
    {
      "epoch": 7.490388621079569,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 31465
    },
    {
      "epoch": 7.490626673808249,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 31466
    },
    {
      "epoch": 7.490864726536928,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 31467
    },
    {
      "epoch": 7.491102779265607,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 31468
    },
    {
      "epoch": 7.491340831994287,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 31469
    },
    {
      "epoch": 7.4915788847229665,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 31470
    },
    {
      "epoch": 7.491816937451645,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 31471
    },
    {
      "epoch": 7.492054990180325,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 31472
    },
    {
      "epoch": 7.492293042909004,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 31473
    },
    {
      "epoch": 7.492531095637684,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 31474
    },
    {
      "epoch": 7.492769148366363,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 31475
    },
    {
      "epoch": 7.493007201095042,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 31476
    },
    {
      "epoch": 7.493245253823722,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 31477
    },
    {
      "epoch": 7.493483306552402,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 31478
    },
    {
      "epoch": 7.49372135928108,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 31479
    },
    {
      "epoch": 7.49395941200976,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0528,
      "step": 31480
    },
    {
      "epoch": 7.4941974647384395,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 31481
    },
    {
      "epoch": 7.494435517467119,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 31482
    },
    {
      "epoch": 7.494673570195799,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 31483
    },
    {
      "epoch": 7.494911622924477,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 31484
    },
    {
      "epoch": 7.495149675653157,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 31485
    },
    {
      "epoch": 7.495387728381837,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 31486
    },
    {
      "epoch": 7.495625781110516,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 31487
    },
    {
      "epoch": 7.495863833839195,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2719,
      "step": 31488
    },
    {
      "epoch": 7.496101886567875,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 31489
    },
    {
      "epoch": 7.496339939296554,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 31490
    },
    {
      "epoch": 7.496577992025234,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 31491
    },
    {
      "epoch": 7.496816044753913,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 31492
    },
    {
      "epoch": 7.497054097482592,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 31493
    },
    {
      "epoch": 7.497292150211272,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 31494
    },
    {
      "epoch": 7.497530202939951,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 31495
    },
    {
      "epoch": 7.497768255668631,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 31496
    },
    {
      "epoch": 7.49800630839731,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 31497
    },
    {
      "epoch": 7.498244361125989,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 31498
    },
    {
      "epoch": 7.498482413854669,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 31499
    },
    {
      "epoch": 7.4987204665833485,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 31500
    },
    {
      "epoch": 7.4987204665833485,
      "eval_runtime": 0.5285,
      "eval_samples_per_second": 1.892,
      "eval_steps_per_second": 1.892,
      "step": 31500
    },
    {
      "epoch": 7.498958519312028,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 31501
    },
    {
      "epoch": 7.499196572040707,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4436,
      "step": 31502
    },
    {
      "epoch": 7.499434624769386,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8002,
      "step": 31503
    },
    {
      "epoch": 7.499672677498066,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 31504
    },
    {
      "epoch": 7.499910730226746,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 31505
    },
    {
      "epoch": 7.500148782955424,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 31506
    },
    {
      "epoch": 7.500386835684104,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 31507
    },
    {
      "epoch": 7.5006248884127835,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 31508
    },
    {
      "epoch": 7.500862941141463,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 31509
    },
    {
      "epoch": 7.501100993870143,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 31510
    },
    {
      "epoch": 7.5013390465988214,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3982,
      "step": 31511
    },
    {
      "epoch": 7.501577099327501,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 31512
    },
    {
      "epoch": 7.501815152056181,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4008,
      "step": 31513
    },
    {
      "epoch": 7.502053204784859,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 31514
    },
    {
      "epoch": 7.502291257513539,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 31515
    },
    {
      "epoch": 7.502529310242219,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 31516
    },
    {
      "epoch": 7.502767362970898,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6131,
      "step": 31517
    },
    {
      "epoch": 7.503005415699578,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4418,
      "step": 31518
    },
    {
      "epoch": 7.5032434684282565,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0413,
      "step": 31519
    },
    {
      "epoch": 7.503481521156936,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 31520
    },
    {
      "epoch": 7.503719573885616,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2734,
      "step": 31521
    },
    {
      "epoch": 7.503957626614295,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8419,
      "step": 31522
    },
    {
      "epoch": 7.504195679342974,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1238,
      "step": 31523
    },
    {
      "epoch": 7.504433732071654,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 31524
    },
    {
      "epoch": 7.504671784800333,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 31525
    },
    {
      "epoch": 7.504909837529013,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 31526
    },
    {
      "epoch": 7.5051478902576925,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 31527
    },
    {
      "epoch": 7.505385942986371,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 31528
    },
    {
      "epoch": 7.505623995715051,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 31529
    },
    {
      "epoch": 7.50586204844373,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3445,
      "step": 31530
    },
    {
      "epoch": 7.50610010117241,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 31531
    },
    {
      "epoch": 7.506338153901089,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 31532
    },
    {
      "epoch": 7.506576206629768,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 31533
    },
    {
      "epoch": 7.506814259358448,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 31534
    },
    {
      "epoch": 7.5070523120871275,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.408,
      "step": 31535
    },
    {
      "epoch": 7.507290364815807,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.5218,
      "step": 31536
    },
    {
      "epoch": 7.507528417544486,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 31537
    },
    {
      "epoch": 7.5077664702731655,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 31538
    },
    {
      "epoch": 7.508004523001845,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 31539
    },
    {
      "epoch": 7.508242575730525,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2139,
      "step": 31540
    },
    {
      "epoch": 7.508480628459203,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 31541
    },
    {
      "epoch": 7.508718681187883,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 31542
    },
    {
      "epoch": 7.508956733916563,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 31543
    },
    {
      "epoch": 7.509194786645242,
      "grad_norm": 6.6875,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 31544
    },
    {
      "epoch": 7.509432839373922,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 31545
    },
    {
      "epoch": 7.5096708921026005,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 31546
    },
    {
      "epoch": 7.50990894483128,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 31547
    },
    {
      "epoch": 7.51014699755996,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2834,
      "step": 31548
    },
    {
      "epoch": 7.5103850502886385,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 31549
    },
    {
      "epoch": 7.510623103017318,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 31550
    },
    {
      "epoch": 7.510861155745998,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3754,
      "step": 31551
    },
    {
      "epoch": 7.511099208474677,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 31552
    },
    {
      "epoch": 7.511337261203357,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8559,
      "step": 31553
    },
    {
      "epoch": 7.511575313932036,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 31554
    },
    {
      "epoch": 7.511813366660715,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 31555
    },
    {
      "epoch": 7.512051419389395,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 31556
    },
    {
      "epoch": 7.512289472118074,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 31557
    },
    {
      "epoch": 7.512527524846753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 31558
    },
    {
      "epoch": 7.512765577575433,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 31559
    },
    {
      "epoch": 7.513003630304112,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 31560
    },
    {
      "epoch": 7.513241683032792,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 31561
    },
    {
      "epoch": 7.5134797357614715,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1343,
      "step": 31562
    },
    {
      "epoch": 7.51371778849015,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 31563
    },
    {
      "epoch": 7.51395584121883,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2131,
      "step": 31564
    },
    {
      "epoch": 7.5141938939475095,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 31565
    },
    {
      "epoch": 7.514431946676189,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 31566
    },
    {
      "epoch": 7.514669999404868,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 31567
    },
    {
      "epoch": 7.514908052133547,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 31568
    },
    {
      "epoch": 7.515146104862227,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 31569
    },
    {
      "epoch": 7.515384157590907,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 31570
    },
    {
      "epoch": 7.515622210319586,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 31571
    },
    {
      "epoch": 7.515860263048265,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 31572
    },
    {
      "epoch": 7.5160983157769445,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 31573
    },
    {
      "epoch": 7.516336368505624,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4116,
      "step": 31574
    },
    {
      "epoch": 7.516574421234304,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 31575
    },
    {
      "epoch": 7.5168124739629825,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3962,
      "step": 31576
    },
    {
      "epoch": 7.517050526691662,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 31577
    },
    {
      "epoch": 7.517288579420342,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8505,
      "step": 31578
    },
    {
      "epoch": 7.517526632149021,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 31579
    },
    {
      "epoch": 7.517764684877701,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 31580
    },
    {
      "epoch": 7.51800273760638,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 31581
    },
    {
      "epoch": 7.518240790335059,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 31582
    },
    {
      "epoch": 7.518478843063739,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 31583
    },
    {
      "epoch": 7.5187168957924175,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 31584
    },
    {
      "epoch": 7.518954948521097,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 31585
    },
    {
      "epoch": 7.519193001249777,
      "grad_norm": 7.03125,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 31586
    },
    {
      "epoch": 7.519431053978456,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 31587
    },
    {
      "epoch": 7.519669106707136,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 31588
    },
    {
      "epoch": 7.519907159435815,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 31589
    },
    {
      "epoch": 7.520145212164494,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.855,
      "step": 31590
    },
    {
      "epoch": 7.520383264893174,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 31591
    },
    {
      "epoch": 7.5206213176218535,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 31592
    },
    {
      "epoch": 7.520859370350532,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2513,
      "step": 31593
    },
    {
      "epoch": 7.521097423079212,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 31594
    },
    {
      "epoch": 7.521335475807891,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 31595
    },
    {
      "epoch": 7.521573528536571,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.6659,
      "step": 31596
    },
    {
      "epoch": 7.521811581265251,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 31597
    },
    {
      "epoch": 7.522049633993929,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2282,
      "step": 31598
    },
    {
      "epoch": 7.522287686722609,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 31599
    },
    {
      "epoch": 7.5225257394512886,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 31600
    },
    {
      "epoch": 7.5225257394512886,
      "eval_runtime": 0.5157,
      "eval_samples_per_second": 1.939,
      "eval_steps_per_second": 1.939,
      "step": 31600
    },
    {
      "epoch": 7.522763792179968,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 31601
    },
    {
      "epoch": 7.523001844908647,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 31602
    },
    {
      "epoch": 7.5232398976373265,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 31603
    },
    {
      "epoch": 7.523477950366006,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 31604
    },
    {
      "epoch": 7.523716003094686,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 31605
    },
    {
      "epoch": 7.523954055823365,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 31606
    },
    {
      "epoch": 7.524192108552044,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 31607
    },
    {
      "epoch": 7.524430161280724,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 31608
    },
    {
      "epoch": 7.524668214009403,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 31609
    },
    {
      "epoch": 7.524906266738083,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 31610
    },
    {
      "epoch": 7.5251443194667615,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9132,
      "step": 31611
    },
    {
      "epoch": 7.525382372195441,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 31612
    },
    {
      "epoch": 7.525620424924121,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 31613
    },
    {
      "epoch": 7.5258584776528,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1875,
      "step": 31614
    },
    {
      "epoch": 7.52609653038148,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 31615
    },
    {
      "epoch": 7.526334583110159,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1253,
      "step": 31616
    },
    {
      "epoch": 7.526572635838838,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.104,
      "step": 31617
    },
    {
      "epoch": 7.526810688567518,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 31618
    },
    {
      "epoch": 7.527048741296197,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 31619
    },
    {
      "epoch": 7.527286794024876,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 31620
    },
    {
      "epoch": 7.527524846753556,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 31621
    },
    {
      "epoch": 7.527762899482235,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 31622
    },
    {
      "epoch": 7.528000952210915,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 31623
    },
    {
      "epoch": 7.528239004939594,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 31624
    },
    {
      "epoch": 7.528477057668273,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 31625
    },
    {
      "epoch": 7.528715110396953,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 31626
    },
    {
      "epoch": 7.528953163125633,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 31627
    },
    {
      "epoch": 7.529191215854311,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 31628
    },
    {
      "epoch": 7.529429268582991,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 31629
    },
    {
      "epoch": 7.5296673213116705,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 31630
    },
    {
      "epoch": 7.52990537404035,
      "grad_norm": 6.90625,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 31631
    },
    {
      "epoch": 7.53014342676903,
      "grad_norm": 6.8125,
      "learning_rate": 3e-05,
      "loss": 2.2121,
      "step": 31632
    },
    {
      "epoch": 7.530381479497708,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 31633
    },
    {
      "epoch": 7.530619532226388,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 31634
    },
    {
      "epoch": 7.530857584955068,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 31635
    },
    {
      "epoch": 7.531095637683747,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 31636
    },
    {
      "epoch": 7.531333690412426,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2312,
      "step": 31637
    },
    {
      "epoch": 7.531571743141106,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 31638
    },
    {
      "epoch": 7.531809795869785,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 31639
    },
    {
      "epoch": 7.532047848598465,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 31640
    },
    {
      "epoch": 7.532285901327144,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 31641
    },
    {
      "epoch": 7.532523954055823,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 31642
    },
    {
      "epoch": 7.532762006784503,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3067,
      "step": 31643
    },
    {
      "epoch": 7.533000059513182,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 31644
    },
    {
      "epoch": 7.533238112241862,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 31645
    },
    {
      "epoch": 7.533476164970541,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 31646
    },
    {
      "epoch": 7.53371421769922,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 31647
    },
    {
      "epoch": 7.5339522704279,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2265,
      "step": 31648
    },
    {
      "epoch": 7.534190323156579,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 31649
    },
    {
      "epoch": 7.534428375885259,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 31650
    },
    {
      "epoch": 7.534666428613938,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 31651
    },
    {
      "epoch": 7.534904481342617,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 31652
    },
    {
      "epoch": 7.535142534071297,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 31653
    },
    {
      "epoch": 7.535380586799977,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 31654
    },
    {
      "epoch": 7.535618639528655,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 31655
    },
    {
      "epoch": 7.535856692257335,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7707,
      "step": 31656
    },
    {
      "epoch": 7.5360947449860145,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 31657
    },
    {
      "epoch": 7.536332797714694,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.892,
      "step": 31658
    },
    {
      "epoch": 7.536570850443373,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 31659
    },
    {
      "epoch": 7.536808903172052,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1694,
      "step": 31660
    },
    {
      "epoch": 7.537046955900732,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 31661
    },
    {
      "epoch": 7.537285008629412,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 31662
    },
    {
      "epoch": 7.53752306135809,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 31663
    },
    {
      "epoch": 7.53776111408677,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 31664
    },
    {
      "epoch": 7.53799916681545,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 31665
    },
    {
      "epoch": 7.538237219544129,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 31666
    },
    {
      "epoch": 7.538475272272809,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 31667
    },
    {
      "epoch": 7.5387133250014875,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 31668
    },
    {
      "epoch": 7.538951377730167,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 31669
    },
    {
      "epoch": 7.539189430458847,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 31670
    },
    {
      "epoch": 7.539427483187526,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7654,
      "step": 31671
    },
    {
      "epoch": 7.539665535916205,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 31672
    },
    {
      "epoch": 7.539903588644885,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 31673
    },
    {
      "epoch": 7.540141641373564,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3483,
      "step": 31674
    },
    {
      "epoch": 7.540379694102244,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 31675
    },
    {
      "epoch": 7.5406177468309235,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 31676
    },
    {
      "epoch": 7.540855799559602,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 31677
    },
    {
      "epoch": 7.541093852288282,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2233,
      "step": 31678
    },
    {
      "epoch": 7.541331905016961,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 31679
    },
    {
      "epoch": 7.541569957745641,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2628,
      "step": 31680
    },
    {
      "epoch": 7.54180801047432,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 31681
    },
    {
      "epoch": 7.542046063202999,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 31682
    },
    {
      "epoch": 7.542284115931679,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1556,
      "step": 31683
    },
    {
      "epoch": 7.5425221686603585,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.353,
      "step": 31684
    },
    {
      "epoch": 7.542760221389038,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 31685
    },
    {
      "epoch": 7.542998274117717,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 31686
    },
    {
      "epoch": 7.5432363268463964,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.4604,
      "step": 31687
    },
    {
      "epoch": 7.543474379575076,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 31688
    },
    {
      "epoch": 7.543712432303756,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4612,
      "step": 31689
    },
    {
      "epoch": 7.543950485032434,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 31690
    },
    {
      "epoch": 7.544188537761114,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 31691
    },
    {
      "epoch": 7.544426590489794,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 31692
    },
    {
      "epoch": 7.544664643218473,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 31693
    },
    {
      "epoch": 7.544902695947153,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7402,
      "step": 31694
    },
    {
      "epoch": 7.5451407486758315,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 31695
    },
    {
      "epoch": 7.545378801404511,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.5819,
      "step": 31696
    },
    {
      "epoch": 7.545616854133191,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 31697
    },
    {
      "epoch": 7.545854906861869,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4759,
      "step": 31698
    },
    {
      "epoch": 7.546092959590549,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 31699
    },
    {
      "epoch": 7.546331012319229,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 31700
    },
    {
      "epoch": 7.546331012319229,
      "eval_runtime": 0.5197,
      "eval_samples_per_second": 1.924,
      "eval_steps_per_second": 1.924,
      "step": 31700
    },
    {
      "epoch": 7.546569065047908,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3833,
      "step": 31701
    },
    {
      "epoch": 7.546807117776588,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 31702
    },
    {
      "epoch": 7.547045170505267,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 31703
    },
    {
      "epoch": 7.547283223233946,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 31704
    },
    {
      "epoch": 7.547521275962626,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 31705
    },
    {
      "epoch": 7.547759328691305,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2862,
      "step": 31706
    },
    {
      "epoch": 7.547997381419984,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 31707
    },
    {
      "epoch": 7.548235434148664,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3033,
      "step": 31708
    },
    {
      "epoch": 7.548473486877343,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 31709
    },
    {
      "epoch": 7.548711539606023,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 31710
    },
    {
      "epoch": 7.5489495923347025,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 31711
    },
    {
      "epoch": 7.549187645063381,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 31712
    },
    {
      "epoch": 7.549425697792061,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 31713
    },
    {
      "epoch": 7.5496637505207405,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7994,
      "step": 31714
    },
    {
      "epoch": 7.54990180324942,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9525,
      "step": 31715
    },
    {
      "epoch": 7.550139855978099,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 31716
    },
    {
      "epoch": 7.550377908706778,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 31717
    },
    {
      "epoch": 7.550615961435458,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2573,
      "step": 31718
    },
    {
      "epoch": 7.550854014164138,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 31719
    },
    {
      "epoch": 7.551092066892817,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 31720
    },
    {
      "epoch": 7.551330119621496,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 31721
    },
    {
      "epoch": 7.5515681723501755,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3314,
      "step": 31722
    },
    {
      "epoch": 7.551806225078855,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 31723
    },
    {
      "epoch": 7.552044277807535,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 31724
    },
    {
      "epoch": 7.5522823305362135,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 31725
    },
    {
      "epoch": 7.552520383264893,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 31726
    },
    {
      "epoch": 7.552758435993573,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 31727
    },
    {
      "epoch": 7.552996488722252,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 31728
    },
    {
      "epoch": 7.553234541450932,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 31729
    },
    {
      "epoch": 7.553472594179611,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 31730
    },
    {
      "epoch": 7.55371064690829,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 31731
    },
    {
      "epoch": 7.55394869963697,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 31732
    },
    {
      "epoch": 7.5541867523656485,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 31733
    },
    {
      "epoch": 7.554424805094328,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.6705,
      "step": 31734
    },
    {
      "epoch": 7.554662857823008,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 31735
    },
    {
      "epoch": 7.554900910551687,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.5588,
      "step": 31736
    },
    {
      "epoch": 7.555138963280367,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 31737
    },
    {
      "epoch": 7.555377016009046,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 31738
    },
    {
      "epoch": 7.555615068737725,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4036,
      "step": 31739
    },
    {
      "epoch": 7.555853121466405,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9246,
      "step": 31740
    },
    {
      "epoch": 7.5560911741950845,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 31741
    },
    {
      "epoch": 7.556329226923763,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 31742
    },
    {
      "epoch": 7.556567279652443,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1261,
      "step": 31743
    },
    {
      "epoch": 7.556805332381122,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 31744
    },
    {
      "epoch": 7.557043385109802,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 31745
    },
    {
      "epoch": 7.557281437838482,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 31746
    },
    {
      "epoch": 7.55751949056716,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 31747
    },
    {
      "epoch": 7.55775754329584,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 31748
    },
    {
      "epoch": 7.5579955960245195,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 31749
    },
    {
      "epoch": 7.558233648753199,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 31750
    },
    {
      "epoch": 7.558471701481878,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 31751
    },
    {
      "epoch": 7.5587097542105575,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 31752
    },
    {
      "epoch": 7.558947806939237,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1746,
      "step": 31753
    },
    {
      "epoch": 7.559185859667917,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 31754
    },
    {
      "epoch": 7.559423912396596,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 31755
    },
    {
      "epoch": 7.559661965125275,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 31756
    },
    {
      "epoch": 7.559900017853955,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 31757
    },
    {
      "epoch": 7.560138070582634,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 31758
    },
    {
      "epoch": 7.560376123311314,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 31759
    },
    {
      "epoch": 7.5606141760399925,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 31760
    },
    {
      "epoch": 7.560852228768672,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 31761
    },
    {
      "epoch": 7.561090281497352,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 31762
    },
    {
      "epoch": 7.561328334226031,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 31763
    },
    {
      "epoch": 7.561566386954711,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 31764
    },
    {
      "epoch": 7.56180443968339,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2342,
      "step": 31765
    },
    {
      "epoch": 7.562042492412069,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 31766
    },
    {
      "epoch": 7.562280545140749,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 31767
    },
    {
      "epoch": 7.562518597869428,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 31768
    },
    {
      "epoch": 7.562756650598107,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.4262,
      "step": 31769
    },
    {
      "epoch": 7.562994703326787,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 31770
    },
    {
      "epoch": 7.563232756055466,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 31771
    },
    {
      "epoch": 7.563470808784146,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 31772
    },
    {
      "epoch": 7.563708861512825,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4748,
      "step": 31773
    },
    {
      "epoch": 7.563946914241504,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 31774
    },
    {
      "epoch": 7.564184966970184,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 31775
    },
    {
      "epoch": 7.5644230196988635,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 31776
    },
    {
      "epoch": 7.564661072427542,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0367,
      "step": 31777
    },
    {
      "epoch": 7.564899125156222,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 31778
    },
    {
      "epoch": 7.5651371778849015,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 31779
    },
    {
      "epoch": 7.565375230613581,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 31780
    },
    {
      "epoch": 7.565613283342261,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 31781
    },
    {
      "epoch": 7.565851336070939,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 31782
    },
    {
      "epoch": 7.566089388799619,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 31783
    },
    {
      "epoch": 7.566327441528299,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 31784
    },
    {
      "epoch": 7.566565494256978,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 31785
    },
    {
      "epoch": 7.566803546985657,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 31786
    },
    {
      "epoch": 7.5670415997143365,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 31787
    },
    {
      "epoch": 7.567279652443016,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 31788
    },
    {
      "epoch": 7.567517705171696,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0109,
      "step": 31789
    },
    {
      "epoch": 7.567755757900375,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3798,
      "step": 31790
    },
    {
      "epoch": 7.567993810629054,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 31791
    },
    {
      "epoch": 7.568231863357734,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 31792
    },
    {
      "epoch": 7.568469916086413,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9297,
      "step": 31793
    },
    {
      "epoch": 7.568707968815093,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0872,
      "step": 31794
    },
    {
      "epoch": 7.568946021543772,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.5859,
      "step": 31795
    },
    {
      "epoch": 7.569184074272451,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 31796
    },
    {
      "epoch": 7.569422127001131,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3612,
      "step": 31797
    },
    {
      "epoch": 7.56966017972981,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8865,
      "step": 31798
    },
    {
      "epoch": 7.56989823245849,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 31799
    },
    {
      "epoch": 7.570136285187169,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 31800
    },
    {
      "epoch": 7.570136285187169,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 31800
    },
    {
      "epoch": 7.570374337915848,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 31801
    },
    {
      "epoch": 7.570612390644528,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 31802
    },
    {
      "epoch": 7.570850443373207,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 31803
    },
    {
      "epoch": 7.571088496101886,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 31804
    },
    {
      "epoch": 7.571326548830566,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3699,
      "step": 31805
    },
    {
      "epoch": 7.5715646015592455,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9917,
      "step": 31806
    },
    {
      "epoch": 7.571802654287925,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 31807
    },
    {
      "epoch": 7.572040707016604,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 31808
    },
    {
      "epoch": 7.572278759745283,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 31809
    },
    {
      "epoch": 7.572516812473963,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1717,
      "step": 31810
    },
    {
      "epoch": 7.572754865202643,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 31811
    },
    {
      "epoch": 7.572992917931321,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2382,
      "step": 31812
    },
    {
      "epoch": 7.573230970660001,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 31813
    },
    {
      "epoch": 7.5734690233886806,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2795,
      "step": 31814
    },
    {
      "epoch": 7.57370707611736,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 31815
    },
    {
      "epoch": 7.57394512884604,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 31816
    },
    {
      "epoch": 7.5741831815747185,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9998,
      "step": 31817
    },
    {
      "epoch": 7.574421234303398,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 31818
    },
    {
      "epoch": 7.574659287032078,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 31819
    },
    {
      "epoch": 7.574897339760757,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 31820
    },
    {
      "epoch": 7.575135392489436,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 31821
    },
    {
      "epoch": 7.575373445218116,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 31822
    },
    {
      "epoch": 7.575611497946795,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 31823
    },
    {
      "epoch": 7.575849550675475,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2657,
      "step": 31824
    },
    {
      "epoch": 7.576087603404154,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 31825
    },
    {
      "epoch": 7.576325656132833,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 31826
    },
    {
      "epoch": 7.576563708861513,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 31827
    },
    {
      "epoch": 7.576801761590192,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7759,
      "step": 31828
    },
    {
      "epoch": 7.577039814318872,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 31829
    },
    {
      "epoch": 7.577277867047551,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 31830
    },
    {
      "epoch": 7.57751591977623,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 31831
    },
    {
      "epoch": 7.57775397250491,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 31832
    },
    {
      "epoch": 7.5779920252335895,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 31833
    },
    {
      "epoch": 7.578230077962269,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 31834
    },
    {
      "epoch": 7.578468130690948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 31835
    },
    {
      "epoch": 7.578706183419627,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2257,
      "step": 31836
    },
    {
      "epoch": 7.578944236148307,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 31837
    },
    {
      "epoch": 7.579182288876987,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8294,
      "step": 31838
    },
    {
      "epoch": 7.579420341605665,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 31839
    },
    {
      "epoch": 7.579658394334345,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8099,
      "step": 31840
    },
    {
      "epoch": 7.579896447063025,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 31841
    },
    {
      "epoch": 7.580134499791704,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 31842
    },
    {
      "epoch": 7.580372552520383,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 31843
    },
    {
      "epoch": 7.5806106052490625,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.863,
      "step": 31844
    },
    {
      "epoch": 7.580848657977742,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8271,
      "step": 31845
    },
    {
      "epoch": 7.581086710706422,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8025,
      "step": 31846
    },
    {
      "epoch": 7.5813247634351,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 31847
    },
    {
      "epoch": 7.58156281616378,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.5061,
      "step": 31848
    },
    {
      "epoch": 7.58180086889246,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 31849
    },
    {
      "epoch": 7.582038921621139,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 31850
    },
    {
      "epoch": 7.582276974349819,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 31851
    },
    {
      "epoch": 7.582515027078498,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 31852
    },
    {
      "epoch": 7.582753079807177,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9308,
      "step": 31853
    },
    {
      "epoch": 7.582991132535857,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 31854
    },
    {
      "epoch": 7.583229185264536,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 31855
    },
    {
      "epoch": 7.583467237993215,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 31856
    },
    {
      "epoch": 7.583705290721895,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 31857
    },
    {
      "epoch": 7.583943343450574,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0939,
      "step": 31858
    },
    {
      "epoch": 7.584181396179254,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 31859
    },
    {
      "epoch": 7.5844194489079335,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 31860
    },
    {
      "epoch": 7.584657501636612,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1712,
      "step": 31861
    },
    {
      "epoch": 7.584895554365292,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 31862
    },
    {
      "epoch": 7.585133607093971,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3398,
      "step": 31863
    },
    {
      "epoch": 7.585371659822651,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 31864
    },
    {
      "epoch": 7.58560971255133,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 31865
    },
    {
      "epoch": 7.585847765280009,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 31866
    },
    {
      "epoch": 7.586085818008689,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 31867
    },
    {
      "epoch": 7.586323870737369,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 31868
    },
    {
      "epoch": 7.586561923466048,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 31869
    },
    {
      "epoch": 7.586799976194727,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 31870
    },
    {
      "epoch": 7.5870380289234065,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9689,
      "step": 31871
    },
    {
      "epoch": 7.587276081652086,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 31872
    },
    {
      "epoch": 7.587514134380766,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 31873
    },
    {
      "epoch": 7.587752187109444,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8681,
      "step": 31874
    },
    {
      "epoch": 7.587990239838124,
      "grad_norm": 2.734375,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 31875
    },
    {
      "epoch": 7.588228292566804,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 31876
    },
    {
      "epoch": 7.588466345295483,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3131,
      "step": 31877
    },
    {
      "epoch": 7.588704398024162,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 31878
    },
    {
      "epoch": 7.588942450752842,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 31879
    },
    {
      "epoch": 7.589180503481521,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 31880
    },
    {
      "epoch": 7.589418556210201,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 31881
    },
    {
      "epoch": 7.5896566089388795,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2337,
      "step": 31882
    },
    {
      "epoch": 7.589894661667559,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 31883
    },
    {
      "epoch": 7.590132714396239,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.5513,
      "step": 31884
    },
    {
      "epoch": 7.590370767124918,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 31885
    },
    {
      "epoch": 7.590608819853598,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 31886
    },
    {
      "epoch": 7.590846872582277,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3246,
      "step": 31887
    },
    {
      "epoch": 7.591084925310956,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 31888
    },
    {
      "epoch": 7.591322978039636,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4278,
      "step": 31889
    },
    {
      "epoch": 7.5915610307683155,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1971,
      "step": 31890
    },
    {
      "epoch": 7.591799083496994,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 31891
    },
    {
      "epoch": 7.592037136225674,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 31892
    },
    {
      "epoch": 7.592275188954353,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 31893
    },
    {
      "epoch": 7.592513241683033,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 31894
    },
    {
      "epoch": 7.592751294411713,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 31895
    },
    {
      "epoch": 7.592989347140391,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2193,
      "step": 31896
    },
    {
      "epoch": 7.593227399869071,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 31897
    },
    {
      "epoch": 7.5934654525977505,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 31898
    },
    {
      "epoch": 7.59370350532643,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 31899
    },
    {
      "epoch": 7.593941558055109,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 31900
    },
    {
      "epoch": 7.593941558055109,
      "eval_runtime": 0.5324,
      "eval_samples_per_second": 1.878,
      "eval_steps_per_second": 1.878,
      "step": 31900
    },
    {
      "epoch": 7.5941796107837884,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4034,
      "step": 31901
    },
    {
      "epoch": 7.594417663512468,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 31902
    },
    {
      "epoch": 7.594655716241148,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 31903
    },
    {
      "epoch": 7.594893768969827,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 31904
    },
    {
      "epoch": 7.595131821698506,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 31905
    },
    {
      "epoch": 7.595369874427186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 31906
    },
    {
      "epoch": 7.595607927155865,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 31907
    },
    {
      "epoch": 7.595845979884545,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9119,
      "step": 31908
    },
    {
      "epoch": 7.5960840326132235,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 31909
    },
    {
      "epoch": 7.596322085341903,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 31910
    },
    {
      "epoch": 7.596560138070583,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 31911
    },
    {
      "epoch": 7.596798190799262,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 31912
    },
    {
      "epoch": 7.597036243527942,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 31913
    },
    {
      "epoch": 7.597274296256621,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 31914
    },
    {
      "epoch": 7.5975123489853,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 31915
    },
    {
      "epoch": 7.59775040171398,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.3721,
      "step": 31916
    },
    {
      "epoch": 7.597988454442659,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.5166,
      "step": 31917
    },
    {
      "epoch": 7.598226507171338,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 31918
    },
    {
      "epoch": 7.598464559900018,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 31919
    },
    {
      "epoch": 7.598702612628697,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 31920
    },
    {
      "epoch": 7.598940665357377,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 31921
    },
    {
      "epoch": 7.599178718086056,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 31922
    },
    {
      "epoch": 7.599416770814735,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 31923
    },
    {
      "epoch": 7.599654823543415,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 31924
    },
    {
      "epoch": 7.5998928762720945,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 31925
    },
    {
      "epoch": 7.600130929000773,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 31926
    },
    {
      "epoch": 7.600368981729453,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 31927
    },
    {
      "epoch": 7.6006070344581325,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 31928
    },
    {
      "epoch": 7.600845087186812,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 31929
    },
    {
      "epoch": 7.601083139915492,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 31930
    },
    {
      "epoch": 7.60132119264417,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 31931
    },
    {
      "epoch": 7.60155924537285,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 31932
    },
    {
      "epoch": 7.60179729810153,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 31933
    },
    {
      "epoch": 7.602035350830209,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0568,
      "step": 31934
    },
    {
      "epoch": 7.602273403558888,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 31935
    },
    {
      "epoch": 7.6025114562875675,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 31936
    },
    {
      "epoch": 7.602749509016247,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2042,
      "step": 31937
    },
    {
      "epoch": 7.602987561744927,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1256,
      "step": 31938
    },
    {
      "epoch": 7.603225614473606,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2814,
      "step": 31939
    },
    {
      "epoch": 7.603463667202285,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 31940
    },
    {
      "epoch": 7.603701719930965,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 31941
    },
    {
      "epoch": 7.603939772659644,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3742,
      "step": 31942
    },
    {
      "epoch": 7.604177825388324,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9533,
      "step": 31943
    },
    {
      "epoch": 7.604415878117003,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 31944
    },
    {
      "epoch": 7.604653930845682,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 31945
    },
    {
      "epoch": 7.604891983574362,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 31946
    },
    {
      "epoch": 7.605130036303041,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9834,
      "step": 31947
    },
    {
      "epoch": 7.605368089031721,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9941,
      "step": 31948
    },
    {
      "epoch": 7.6056061417604,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 31949
    },
    {
      "epoch": 7.605844194489079,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 31950
    },
    {
      "epoch": 7.606082247217759,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.352,
      "step": 31951
    },
    {
      "epoch": 7.606320299946438,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 31952
    },
    {
      "epoch": 7.606558352675117,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8494,
      "step": 31953
    },
    {
      "epoch": 7.606796405403797,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 31954
    },
    {
      "epoch": 7.6070344581324765,
      "grad_norm": 8.0,
      "learning_rate": 3e-05,
      "loss": 2.2173,
      "step": 31955
    },
    {
      "epoch": 7.607272510861156,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 31956
    },
    {
      "epoch": 7.607510563589835,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.942,
      "step": 31957
    },
    {
      "epoch": 7.607748616318514,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 31958
    },
    {
      "epoch": 7.607986669047194,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 31959
    },
    {
      "epoch": 7.608224721775874,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.358,
      "step": 31960
    },
    {
      "epoch": 7.608462774504552,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4098,
      "step": 31961
    },
    {
      "epoch": 7.608700827233232,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 31962
    },
    {
      "epoch": 7.6089388799619115,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 31963
    },
    {
      "epoch": 7.609176932690591,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2,
      "step": 31964
    },
    {
      "epoch": 7.609414985419271,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8212,
      "step": 31965
    },
    {
      "epoch": 7.6096530381479495,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 31966
    },
    {
      "epoch": 7.609891090876629,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 31967
    },
    {
      "epoch": 7.610129143605309,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 31968
    },
    {
      "epoch": 7.610367196333988,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 31969
    },
    {
      "epoch": 7.610605249062667,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 31970
    },
    {
      "epoch": 7.610843301791347,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 31971
    },
    {
      "epoch": 7.611081354520026,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 31972
    },
    {
      "epoch": 7.611319407248706,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 31973
    },
    {
      "epoch": 7.611557459977385,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 31974
    },
    {
      "epoch": 7.611795512706064,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 31975
    },
    {
      "epoch": 7.612033565434744,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9833,
      "step": 31976
    },
    {
      "epoch": 7.612271618163423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 31977
    },
    {
      "epoch": 7.612509670892103,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 31978
    },
    {
      "epoch": 7.612747723620782,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 31979
    },
    {
      "epoch": 7.612985776349461,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 31980
    },
    {
      "epoch": 7.613223829078141,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2948,
      "step": 31981
    },
    {
      "epoch": 7.6134618818068205,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 31982
    },
    {
      "epoch": 7.6136999345355,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1854,
      "step": 31983
    },
    {
      "epoch": 7.613937987264179,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 31984
    },
    {
      "epoch": 7.614176039992858,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8252,
      "step": 31985
    },
    {
      "epoch": 7.614414092721538,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 31986
    },
    {
      "epoch": 7.614652145450217,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 31987
    },
    {
      "epoch": 7.614890198178896,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 31988
    },
    {
      "epoch": 7.615128250907576,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 31989
    },
    {
      "epoch": 7.6153663036362556,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8663,
      "step": 31990
    },
    {
      "epoch": 7.615604356364935,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 31991
    },
    {
      "epoch": 7.615842409093614,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 31992
    },
    {
      "epoch": 7.6160804618222935,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 31993
    },
    {
      "epoch": 7.616318514550973,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 31994
    },
    {
      "epoch": 7.616556567279653,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 31995
    },
    {
      "epoch": 7.616794620008331,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 31996
    },
    {
      "epoch": 7.617032672737011,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 31997
    },
    {
      "epoch": 7.617270725465691,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 31998
    },
    {
      "epoch": 7.61750877819437,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 31999
    },
    {
      "epoch": 7.61774683092305,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 32000
    },
    {
      "epoch": 7.61774683092305,
      "eval_runtime": 0.5271,
      "eval_samples_per_second": 1.897,
      "eval_steps_per_second": 1.897,
      "step": 32000
    },
    {
      "epoch": 7.6179848836517285,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 32001
    },
    {
      "epoch": 7.618222936380408,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 32002
    },
    {
      "epoch": 7.618460989109088,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1896,
      "step": 32003
    },
    {
      "epoch": 7.618699041837767,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.895,
      "step": 32004
    },
    {
      "epoch": 7.618937094566446,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 32005
    },
    {
      "epoch": 7.619175147295126,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 32006
    },
    {
      "epoch": 7.619413200023805,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 32007
    },
    {
      "epoch": 7.619651252752485,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 32008
    },
    {
      "epoch": 7.6198893054811645,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.79,
      "step": 32009
    },
    {
      "epoch": 7.620127358209843,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 32010
    },
    {
      "epoch": 7.620365410938523,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 32011
    },
    {
      "epoch": 7.620603463667202,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2865,
      "step": 32012
    },
    {
      "epoch": 7.620841516395882,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 32013
    },
    {
      "epoch": 7.621079569124561,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 32014
    },
    {
      "epoch": 7.62131762185324,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2058,
      "step": 32015
    },
    {
      "epoch": 7.62155567458192,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 32016
    },
    {
      "epoch": 7.6217937273106,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 32017
    },
    {
      "epoch": 7.622031780039279,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 32018
    },
    {
      "epoch": 7.622269832767958,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 32019
    },
    {
      "epoch": 7.6225078854966375,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 32020
    },
    {
      "epoch": 7.622745938225317,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 32021
    },
    {
      "epoch": 7.622983990953996,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 32022
    },
    {
      "epoch": 7.623222043682675,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 32023
    },
    {
      "epoch": 7.623460096411355,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 32024
    },
    {
      "epoch": 7.623698149140035,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 32025
    },
    {
      "epoch": 7.623936201868714,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1706,
      "step": 32026
    },
    {
      "epoch": 7.624174254597393,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 32027
    },
    {
      "epoch": 7.624412307326073,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 32028
    },
    {
      "epoch": 7.624650360054752,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 32029
    },
    {
      "epoch": 7.624888412783432,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 32030
    },
    {
      "epoch": 7.6251264655121105,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 32031
    },
    {
      "epoch": 7.62536451824079,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 32032
    },
    {
      "epoch": 7.62560257096947,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 32033
    },
    {
      "epoch": 7.625840623698149,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 32034
    },
    {
      "epoch": 7.626078676426829,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 32035
    },
    {
      "epoch": 7.626316729155508,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 32036
    },
    {
      "epoch": 7.626554781884187,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 32037
    },
    {
      "epoch": 7.626792834612867,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2302,
      "step": 32038
    },
    {
      "epoch": 7.627030887341546,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 32039
    },
    {
      "epoch": 7.627268940070225,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 32040
    },
    {
      "epoch": 7.627506992798905,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7974,
      "step": 32041
    },
    {
      "epoch": 7.627745045527584,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 32042
    },
    {
      "epoch": 7.627983098256264,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.842,
      "step": 32043
    },
    {
      "epoch": 7.628221150984944,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 32044
    },
    {
      "epoch": 7.628459203713622,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 32045
    },
    {
      "epoch": 7.628697256442302,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 32046
    },
    {
      "epoch": 7.6289353091709815,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 32047
    },
    {
      "epoch": 7.629173361899661,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 32048
    },
    {
      "epoch": 7.62941141462834,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 32049
    },
    {
      "epoch": 7.629649467357019,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 32050
    },
    {
      "epoch": 7.629887520085699,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.6286,
      "step": 32051
    },
    {
      "epoch": 7.630125572814379,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 32052
    },
    {
      "epoch": 7.630363625543058,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 32053
    },
    {
      "epoch": 7.630601678271737,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 32054
    },
    {
      "epoch": 7.630839731000417,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 32055
    },
    {
      "epoch": 7.631077783729096,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 32056
    },
    {
      "epoch": 7.631315836457776,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 32057
    },
    {
      "epoch": 7.6315538891864545,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1938,
      "step": 32058
    },
    {
      "epoch": 7.631791941915134,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 32059
    },
    {
      "epoch": 7.632029994643814,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 32060
    },
    {
      "epoch": 7.632268047372493,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 32061
    },
    {
      "epoch": 7.632506100101172,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 32062
    },
    {
      "epoch": 7.632744152829852,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 32063
    },
    {
      "epoch": 7.632982205558531,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 32064
    },
    {
      "epoch": 7.633220258287211,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 32065
    },
    {
      "epoch": 7.63345831101589,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 32066
    },
    {
      "epoch": 7.633696363744569,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 32067
    },
    {
      "epoch": 7.633934416473249,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 32068
    },
    {
      "epoch": 7.634172469201928,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.5149,
      "step": 32069
    },
    {
      "epoch": 7.634410521930608,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 32070
    },
    {
      "epoch": 7.634648574659287,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1565,
      "step": 32071
    },
    {
      "epoch": 7.634886627387966,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 32072
    },
    {
      "epoch": 7.635124680116646,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 32073
    },
    {
      "epoch": 7.6353627328453255,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 32074
    },
    {
      "epoch": 7.635600785574004,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 32075
    },
    {
      "epoch": 7.635838838302684,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 32076
    },
    {
      "epoch": 7.6360768910313634,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0475,
      "step": 32077
    },
    {
      "epoch": 7.636314943760043,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 32078
    },
    {
      "epoch": 7.636552996488723,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 32079
    },
    {
      "epoch": 7.636791049217401,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 32080
    },
    {
      "epoch": 7.637029101946081,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 32081
    },
    {
      "epoch": 7.637267154674761,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 32082
    },
    {
      "epoch": 7.63750520740344,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 32083
    },
    {
      "epoch": 7.637743260132119,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0382,
      "step": 32084
    },
    {
      "epoch": 7.6379813128607985,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9756,
      "step": 32085
    },
    {
      "epoch": 7.638219365589478,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4807,
      "step": 32086
    },
    {
      "epoch": 7.638457418318158,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 32087
    },
    {
      "epoch": 7.638695471046837,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4474,
      "step": 32088
    },
    {
      "epoch": 7.638933523775516,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 32089
    },
    {
      "epoch": 7.639171576504196,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 32090
    },
    {
      "epoch": 7.639409629232875,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 32091
    },
    {
      "epoch": 7.639647681961555,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 32092
    },
    {
      "epoch": 7.639885734690234,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3512,
      "step": 32093
    },
    {
      "epoch": 7.640123787418913,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 32094
    },
    {
      "epoch": 7.640361840147593,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 32095
    },
    {
      "epoch": 7.640599892876272,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 32096
    },
    {
      "epoch": 7.640837945604951,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 32097
    },
    {
      "epoch": 7.641075998333631,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 32098
    },
    {
      "epoch": 7.64131405106231,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 32099
    },
    {
      "epoch": 7.64155210379099,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 32100
    },
    {
      "epoch": 7.64155210379099,
      "eval_runtime": 0.5227,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 32100
    },
    {
      "epoch": 7.641790156519669,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2962,
      "step": 32101
    },
    {
      "epoch": 7.642028209248348,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7848,
      "step": 32102
    },
    {
      "epoch": 7.642266261977028,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 32103
    },
    {
      "epoch": 7.6425043147057075,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 32104
    },
    {
      "epoch": 7.642742367434387,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 32105
    },
    {
      "epoch": 7.642980420163066,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 32106
    },
    {
      "epoch": 7.643218472891745,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0374,
      "step": 32107
    },
    {
      "epoch": 7.643456525620425,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1462,
      "step": 32108
    },
    {
      "epoch": 7.643694578349105,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 32109
    },
    {
      "epoch": 7.643932631077783,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 32110
    },
    {
      "epoch": 7.644170683806463,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8997,
      "step": 32111
    },
    {
      "epoch": 7.6444087365351425,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 32112
    },
    {
      "epoch": 7.644646789263822,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 32113
    },
    {
      "epoch": 7.644884841992502,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.442,
      "step": 32114
    },
    {
      "epoch": 7.6451228947211805,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1271,
      "step": 32115
    },
    {
      "epoch": 7.64536094744986,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 32116
    },
    {
      "epoch": 7.64559900017854,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 32117
    },
    {
      "epoch": 7.645837052907219,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 32118
    },
    {
      "epoch": 7.646075105635898,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 32119
    },
    {
      "epoch": 7.646313158364578,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 32120
    },
    {
      "epoch": 7.646551211093257,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 32121
    },
    {
      "epoch": 7.646789263821937,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 32122
    },
    {
      "epoch": 7.647027316550616,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 32123
    },
    {
      "epoch": 7.647265369279295,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2718,
      "step": 32124
    },
    {
      "epoch": 7.647503422007975,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 32125
    },
    {
      "epoch": 7.647741474736654,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 32126
    },
    {
      "epoch": 7.647979527465334,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 32127
    },
    {
      "epoch": 7.648217580194013,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 32128
    },
    {
      "epoch": 7.648455632922692,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 32129
    },
    {
      "epoch": 7.648693685651372,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 32130
    },
    {
      "epoch": 7.6489317383800515,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 32131
    },
    {
      "epoch": 7.649169791108731,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1703,
      "step": 32132
    },
    {
      "epoch": 7.64940784383741,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.963,
      "step": 32133
    },
    {
      "epoch": 7.649645896566089,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 32134
    },
    {
      "epoch": 7.649883949294769,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 32135
    },
    {
      "epoch": 7.650122002023448,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 32136
    },
    {
      "epoch": 7.650360054752127,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 32137
    },
    {
      "epoch": 7.650598107480807,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 32138
    },
    {
      "epoch": 7.6508361602094865,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 32139
    },
    {
      "epoch": 7.651074212938166,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 32140
    },
    {
      "epoch": 7.651312265666845,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 32141
    },
    {
      "epoch": 7.6515503183955245,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2839,
      "step": 32142
    },
    {
      "epoch": 7.651788371124204,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 32143
    },
    {
      "epoch": 7.652026423852884,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 32144
    },
    {
      "epoch": 7.652264476581562,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 32145
    },
    {
      "epoch": 7.652502529310242,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8846,
      "step": 32146
    },
    {
      "epoch": 7.652740582038922,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 32147
    },
    {
      "epoch": 7.652978634767601,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 32148
    },
    {
      "epoch": 7.653216687496281,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 32149
    },
    {
      "epoch": 7.6534547402249595,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 32150
    },
    {
      "epoch": 7.653692792953639,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.7359,
      "step": 32151
    },
    {
      "epoch": 7.653930845682319,
      "grad_norm": 6.71875,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 32152
    },
    {
      "epoch": 7.654168898410998,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2287,
      "step": 32153
    },
    {
      "epoch": 7.654406951139677,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 32154
    },
    {
      "epoch": 7.654645003868357,
      "grad_norm": 6.59375,
      "learning_rate": 3e-05,
      "loss": 1.7886,
      "step": 32155
    },
    {
      "epoch": 7.654883056597036,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2778,
      "step": 32156
    },
    {
      "epoch": 7.655121109325716,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 32157
    },
    {
      "epoch": 7.6553591620543955,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 32158
    },
    {
      "epoch": 7.655597214783074,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 32159
    },
    {
      "epoch": 7.655835267511754,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 32160
    },
    {
      "epoch": 7.656073320240433,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2489,
      "step": 32161
    },
    {
      "epoch": 7.656311372969113,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 32162
    },
    {
      "epoch": 7.656549425697792,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 32163
    },
    {
      "epoch": 7.656787478426471,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 32164
    },
    {
      "epoch": 7.657025531155151,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 32165
    },
    {
      "epoch": 7.6572635838838305,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1394,
      "step": 32166
    },
    {
      "epoch": 7.65750163661251,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 32167
    },
    {
      "epoch": 7.657739689341189,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 32168
    },
    {
      "epoch": 7.6579777420698685,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.015,
      "step": 32169
    },
    {
      "epoch": 7.658215794798548,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 32170
    },
    {
      "epoch": 7.658453847527227,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 32171
    },
    {
      "epoch": 7.658691900255906,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 32172
    },
    {
      "epoch": 7.658929952984586,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.859,
      "step": 32173
    },
    {
      "epoch": 7.659168005713266,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 32174
    },
    {
      "epoch": 7.659406058441945,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 32175
    },
    {
      "epoch": 7.659644111170624,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 32176
    },
    {
      "epoch": 7.6598821638993035,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1937,
      "step": 32177
    },
    {
      "epoch": 7.660120216627983,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 1.6881,
      "step": 32178
    },
    {
      "epoch": 7.660358269356663,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 32179
    },
    {
      "epoch": 7.6605963220853415,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 32180
    },
    {
      "epoch": 7.660834374814021,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 32181
    },
    {
      "epoch": 7.661072427542701,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 32182
    },
    {
      "epoch": 7.66131048027138,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 32183
    },
    {
      "epoch": 7.66154853300006,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 32184
    },
    {
      "epoch": 7.661786585728739,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.373,
      "step": 32185
    },
    {
      "epoch": 7.662024638457418,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 32186
    },
    {
      "epoch": 7.662262691186098,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 32187
    },
    {
      "epoch": 7.662500743914777,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 32188
    },
    {
      "epoch": 7.662738796643456,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 32189
    },
    {
      "epoch": 7.662976849372136,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4498,
      "step": 32190
    },
    {
      "epoch": 7.663214902100815,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2084,
      "step": 32191
    },
    {
      "epoch": 7.663452954829495,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8444,
      "step": 32192
    },
    {
      "epoch": 7.663691007558175,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2262,
      "step": 32193
    },
    {
      "epoch": 7.663929060286853,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 32194
    },
    {
      "epoch": 7.664167113015533,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8152,
      "step": 32195
    },
    {
      "epoch": 7.6644051657442125,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 32196
    },
    {
      "epoch": 7.664643218472892,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.826,
      "step": 32197
    },
    {
      "epoch": 7.664881271201571,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 32198
    },
    {
      "epoch": 7.66511932393025,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 32199
    },
    {
      "epoch": 7.66535737665893,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 32200
    },
    {
      "epoch": 7.66535737665893,
      "eval_runtime": 0.5333,
      "eval_samples_per_second": 1.875,
      "eval_steps_per_second": 1.875,
      "step": 32200
    },
    {
      "epoch": 7.66559542938761,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 32201
    },
    {
      "epoch": 7.665833482116289,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8987,
      "step": 32202
    },
    {
      "epoch": 7.666071534844968,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 32203
    },
    {
      "epoch": 7.6663095875736476,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 32204
    },
    {
      "epoch": 7.666547640302327,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 32205
    },
    {
      "epoch": 7.666785693031006,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 32206
    },
    {
      "epoch": 7.6670237457596855,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 32207
    },
    {
      "epoch": 7.667261798488365,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 32208
    },
    {
      "epoch": 7.667499851217045,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8413,
      "step": 32209
    },
    {
      "epoch": 7.667737903945724,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 32210
    },
    {
      "epoch": 7.667975956674403,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9203,
      "step": 32211
    },
    {
      "epoch": 7.668214009403083,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 32212
    },
    {
      "epoch": 7.668452062131762,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 32213
    },
    {
      "epoch": 7.668690114860442,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 32214
    },
    {
      "epoch": 7.6689281675891205,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 32215
    },
    {
      "epoch": 7.6691662203178,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 32216
    },
    {
      "epoch": 7.66940427304648,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3457,
      "step": 32217
    },
    {
      "epoch": 7.669642325775159,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 32218
    },
    {
      "epoch": 7.669880378503839,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9215,
      "step": 32219
    },
    {
      "epoch": 7.670118431232518,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 32220
    },
    {
      "epoch": 7.670356483961197,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 32221
    },
    {
      "epoch": 7.670594536689877,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 32222
    },
    {
      "epoch": 7.6708325894185565,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8304,
      "step": 32223
    },
    {
      "epoch": 7.671070642147235,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9046,
      "step": 32224
    },
    {
      "epoch": 7.671308694875915,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 32225
    },
    {
      "epoch": 7.671546747604594,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 32226
    },
    {
      "epoch": 7.671784800333274,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.902,
      "step": 32227
    },
    {
      "epoch": 7.672022853061954,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 32228
    },
    {
      "epoch": 7.672260905790632,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 32229
    },
    {
      "epoch": 7.672498958519312,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 32230
    },
    {
      "epoch": 7.672737011247992,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 32231
    },
    {
      "epoch": 7.672975063976671,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8598,
      "step": 32232
    },
    {
      "epoch": 7.67321311670535,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4032,
      "step": 32233
    },
    {
      "epoch": 7.6734511694340295,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8042,
      "step": 32234
    },
    {
      "epoch": 7.673689222162709,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 32235
    },
    {
      "epoch": 7.673927274891389,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 32236
    },
    {
      "epoch": 7.674165327620068,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 32237
    },
    {
      "epoch": 7.674403380348747,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 32238
    },
    {
      "epoch": 7.674641433077427,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 32239
    },
    {
      "epoch": 7.674879485806106,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0683,
      "step": 32240
    },
    {
      "epoch": 7.675117538534785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0958,
      "step": 32241
    },
    {
      "epoch": 7.675355591263465,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4469,
      "step": 32242
    },
    {
      "epoch": 7.675593643992144,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 32243
    },
    {
      "epoch": 7.675831696720824,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 32244
    },
    {
      "epoch": 7.676069749449503,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 32245
    },
    {
      "epoch": 7.676307802178182,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 32246
    },
    {
      "epoch": 7.676545854906862,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.193,
      "step": 32247
    },
    {
      "epoch": 7.676783907635541,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 32248
    },
    {
      "epoch": 7.677021960364221,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 32249
    },
    {
      "epoch": 7.6772600130929,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 32250
    },
    {
      "epoch": 7.677498065821579,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 32251
    },
    {
      "epoch": 7.677736118550259,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 32252
    },
    {
      "epoch": 7.677974171278938,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9079,
      "step": 32253
    },
    {
      "epoch": 7.678212224007618,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 32254
    },
    {
      "epoch": 7.678450276736297,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 32255
    },
    {
      "epoch": 7.678688329464976,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7997,
      "step": 32256
    },
    {
      "epoch": 7.678926382193656,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7961,
      "step": 32257
    },
    {
      "epoch": 7.679164434922336,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 32258
    },
    {
      "epoch": 7.679402487651014,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 32259
    },
    {
      "epoch": 7.679640540379694,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 32260
    },
    {
      "epoch": 7.6798785931083735,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 32261
    },
    {
      "epoch": 7.680116645837053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 32262
    },
    {
      "epoch": 7.680354698565733,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 32263
    },
    {
      "epoch": 7.680592751294411,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 32264
    },
    {
      "epoch": 7.680830804023091,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 32265
    },
    {
      "epoch": 7.681068856751771,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8457,
      "step": 32266
    },
    {
      "epoch": 7.68130690948045,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 32267
    },
    {
      "epoch": 7.681544962209129,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 32268
    },
    {
      "epoch": 7.681783014937809,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 32269
    },
    {
      "epoch": 7.682021067666488,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 32270
    },
    {
      "epoch": 7.682259120395168,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 32271
    },
    {
      "epoch": 7.682497173123847,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 32272
    },
    {
      "epoch": 7.682735225852526,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2878,
      "step": 32273
    },
    {
      "epoch": 7.682973278581206,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 32274
    },
    {
      "epoch": 7.683211331309885,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.6445,
      "step": 32275
    },
    {
      "epoch": 7.683449384038565,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 32276
    },
    {
      "epoch": 7.683687436767244,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 32277
    },
    {
      "epoch": 7.683925489495923,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 32278
    },
    {
      "epoch": 7.684163542224603,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 32279
    },
    {
      "epoch": 7.6844015949532825,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 32280
    },
    {
      "epoch": 7.684639647681961,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 32281
    },
    {
      "epoch": 7.684877700410641,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 32282
    },
    {
      "epoch": 7.68511575313932,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 32283
    },
    {
      "epoch": 7.685353805868,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 32284
    },
    {
      "epoch": 7.685591858596679,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.351,
      "step": 32285
    },
    {
      "epoch": 7.685829911325358,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 32286
    },
    {
      "epoch": 7.686067964054038,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 32287
    },
    {
      "epoch": 7.6863060167827175,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 32288
    },
    {
      "epoch": 7.686544069511397,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 32289
    },
    {
      "epoch": 7.686782122240076,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 32290
    },
    {
      "epoch": 7.6870201749687554,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 32291
    },
    {
      "epoch": 7.687258227697435,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 32292
    },
    {
      "epoch": 7.687496280426115,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 32293
    },
    {
      "epoch": 7.687734333154793,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 32294
    },
    {
      "epoch": 7.687972385883473,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 32295
    },
    {
      "epoch": 7.688210438612153,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3249,
      "step": 32296
    },
    {
      "epoch": 7.688448491340832,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.6231,
      "step": 32297
    },
    {
      "epoch": 7.688686544069512,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 32298
    },
    {
      "epoch": 7.6889245967981905,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 32299
    },
    {
      "epoch": 7.68916264952687,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.162,
      "step": 32300
    },
    {
      "epoch": 7.68916264952687,
      "eval_runtime": 0.508,
      "eval_samples_per_second": 1.968,
      "eval_steps_per_second": 1.968,
      "step": 32300
    },
    {
      "epoch": 7.68940070225555,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 32301
    },
    {
      "epoch": 7.689638754984229,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 32302
    },
    {
      "epoch": 7.689876807712908,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2859,
      "step": 32303
    },
    {
      "epoch": 7.690114860441588,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 32304
    },
    {
      "epoch": 7.690352913170267,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2867,
      "step": 32305
    },
    {
      "epoch": 7.690590965898947,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8373,
      "step": 32306
    },
    {
      "epoch": 7.6908290186276265,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 32307
    },
    {
      "epoch": 7.691067071356305,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 32308
    },
    {
      "epoch": 7.691305124084985,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 32309
    },
    {
      "epoch": 7.691543176813664,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 32310
    },
    {
      "epoch": 7.691781229542344,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 32311
    },
    {
      "epoch": 7.692019282271023,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 32312
    },
    {
      "epoch": 7.692257334999702,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2568,
      "step": 32313
    },
    {
      "epoch": 7.692495387728382,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 32314
    },
    {
      "epoch": 7.6927334404570615,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2861,
      "step": 32315
    },
    {
      "epoch": 7.692971493185741,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0016,
      "step": 32316
    },
    {
      "epoch": 7.69320954591442,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8498,
      "step": 32317
    },
    {
      "epoch": 7.6934475986430995,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 32318
    },
    {
      "epoch": 7.693685651371779,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 32319
    },
    {
      "epoch": 7.693923704100458,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 32320
    },
    {
      "epoch": 7.694161756829137,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3502,
      "step": 32321
    },
    {
      "epoch": 7.694399809557817,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 32322
    },
    {
      "epoch": 7.694637862286497,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 32323
    },
    {
      "epoch": 7.694875915015176,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 32324
    },
    {
      "epoch": 7.695113967743855,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8538,
      "step": 32325
    },
    {
      "epoch": 7.6953520204725345,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.418,
      "step": 32326
    },
    {
      "epoch": 7.695590073201214,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 32327
    },
    {
      "epoch": 7.695828125929894,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7934,
      "step": 32328
    },
    {
      "epoch": 7.6960661786585725,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 32329
    },
    {
      "epoch": 7.696304231387252,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 32330
    },
    {
      "epoch": 7.696542284115932,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 32331
    },
    {
      "epoch": 7.696780336844611,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 32332
    },
    {
      "epoch": 7.697018389573291,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 32333
    },
    {
      "epoch": 7.69725644230197,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 32334
    },
    {
      "epoch": 7.697494495030649,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 32335
    },
    {
      "epoch": 7.697732547759329,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 32336
    },
    {
      "epoch": 7.697970600488008,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 32337
    },
    {
      "epoch": 7.698208653216687,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.5045,
      "step": 32338
    },
    {
      "epoch": 7.698446705945367,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 32339
    },
    {
      "epoch": 7.698684758674046,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3862,
      "step": 32340
    },
    {
      "epoch": 7.698922811402726,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 32341
    },
    {
      "epoch": 7.6991608641314055,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9757,
      "step": 32342
    },
    {
      "epoch": 7.699398916860084,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 32343
    },
    {
      "epoch": 7.699636969588764,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 32344
    },
    {
      "epoch": 7.6998750223174435,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8492,
      "step": 32345
    },
    {
      "epoch": 7.700113075046123,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 32346
    },
    {
      "epoch": 7.700351127774802,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 32347
    },
    {
      "epoch": 7.700589180503481,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 32348
    },
    {
      "epoch": 7.700827233232161,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3024,
      "step": 32349
    },
    {
      "epoch": 7.701065285960841,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 32350
    },
    {
      "epoch": 7.70130333868952,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.3084,
      "step": 32351
    },
    {
      "epoch": 7.701541391418199,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 32352
    },
    {
      "epoch": 7.7017794441468785,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 32353
    },
    {
      "epoch": 7.702017496875558,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 32354
    },
    {
      "epoch": 7.702255549604237,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 32355
    },
    {
      "epoch": 7.7024936023329165,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 32356
    },
    {
      "epoch": 7.702731655061596,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 32357
    },
    {
      "epoch": 7.702969707790276,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2408,
      "step": 32358
    },
    {
      "epoch": 7.703207760518955,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 32359
    },
    {
      "epoch": 7.703445813247634,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 32360
    },
    {
      "epoch": 7.703683865976314,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 32361
    },
    {
      "epoch": 7.703921918704993,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 32362
    },
    {
      "epoch": 7.704159971433673,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9349,
      "step": 32363
    },
    {
      "epoch": 7.7043980241623515,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 32364
    },
    {
      "epoch": 7.704636076891031,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 32365
    },
    {
      "epoch": 7.704874129619711,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.239,
      "step": 32366
    },
    {
      "epoch": 7.70511218234839,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 32367
    },
    {
      "epoch": 7.70535023507707,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 32368
    },
    {
      "epoch": 7.705588287805749,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 32369
    },
    {
      "epoch": 7.705826340534428,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 32370
    },
    {
      "epoch": 7.706064393263108,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 32371
    },
    {
      "epoch": 7.7063024459917875,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 32372
    },
    {
      "epoch": 7.706540498720466,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 32373
    },
    {
      "epoch": 7.706778551449146,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.0948,
      "step": 32374
    },
    {
      "epoch": 7.707016604177825,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 32375
    },
    {
      "epoch": 7.707254656906505,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 32376
    },
    {
      "epoch": 7.707492709635185,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 32377
    },
    {
      "epoch": 7.707730762363863,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 32378
    },
    {
      "epoch": 7.707968815092543,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 32379
    },
    {
      "epoch": 7.7082068678212226,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 32380
    },
    {
      "epoch": 7.708444920549902,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 32381
    },
    {
      "epoch": 7.708682973278581,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 32382
    },
    {
      "epoch": 7.7089210260072605,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9225,
      "step": 32383
    },
    {
      "epoch": 7.70915907873594,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 32384
    },
    {
      "epoch": 7.70939713146462,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 32385
    },
    {
      "epoch": 7.709635184193299,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3287,
      "step": 32386
    },
    {
      "epoch": 7.709873236921978,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 32387
    },
    {
      "epoch": 7.710111289650658,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3238,
      "step": 32388
    },
    {
      "epoch": 7.710349342379337,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 32389
    },
    {
      "epoch": 7.710587395108016,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 32390
    },
    {
      "epoch": 7.7108254478366955,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 32391
    },
    {
      "epoch": 7.711063500565375,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 32392
    },
    {
      "epoch": 7.711301553294055,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7474,
      "step": 32393
    },
    {
      "epoch": 7.711539606022734,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1998,
      "step": 32394
    },
    {
      "epoch": 7.711777658751413,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1615,
      "step": 32395
    },
    {
      "epoch": 7.712015711480093,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 32396
    },
    {
      "epoch": 7.712253764208772,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 32397
    },
    {
      "epoch": 7.712491816937452,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 32398
    },
    {
      "epoch": 7.712729869666131,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1207,
      "step": 32399
    },
    {
      "epoch": 7.71296792239481,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 32400
    },
    {
      "epoch": 7.71296792239481,
      "eval_runtime": 0.5245,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 32400
    },
    {
      "epoch": 7.71320597512349,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 32401
    },
    {
      "epoch": 7.713444027852169,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7568,
      "step": 32402
    },
    {
      "epoch": 7.713682080580849,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8571,
      "step": 32403
    },
    {
      "epoch": 7.713920133309528,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9892,
      "step": 32404
    },
    {
      "epoch": 7.714158186038207,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 32405
    },
    {
      "epoch": 7.714396238766887,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 32406
    },
    {
      "epoch": 7.714634291495567,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 32407
    },
    {
      "epoch": 7.714872344224245,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1345,
      "step": 32408
    },
    {
      "epoch": 7.715110396952925,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1667,
      "step": 32409
    },
    {
      "epoch": 7.7153484496816045,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 32410
    },
    {
      "epoch": 7.715586502410284,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 32411
    },
    {
      "epoch": 7.715824555138964,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 32412
    },
    {
      "epoch": 7.716062607867642,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 32413
    },
    {
      "epoch": 7.716300660596322,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 32414
    },
    {
      "epoch": 7.716538713325002,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 32415
    },
    {
      "epoch": 7.716776766053681,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 32416
    },
    {
      "epoch": 7.71701481878236,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 32417
    },
    {
      "epoch": 7.71725287151104,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2803,
      "step": 32418
    },
    {
      "epoch": 7.717490924239719,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 32419
    },
    {
      "epoch": 7.717728976968399,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3191,
      "step": 32420
    },
    {
      "epoch": 7.717967029697078,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 32421
    },
    {
      "epoch": 7.718205082425757,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 32422
    },
    {
      "epoch": 7.718443135154437,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 32423
    },
    {
      "epoch": 7.718681187883116,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 32424
    },
    {
      "epoch": 7.718919240611795,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 32425
    },
    {
      "epoch": 7.719157293340475,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 32426
    },
    {
      "epoch": 7.719395346069154,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 32427
    },
    {
      "epoch": 7.719633398797834,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 32428
    },
    {
      "epoch": 7.719871451526513,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 32429
    },
    {
      "epoch": 7.720109504255192,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 32430
    },
    {
      "epoch": 7.720347556983872,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 32431
    },
    {
      "epoch": 7.720585609712551,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 32432
    },
    {
      "epoch": 7.720823662441231,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 32433
    },
    {
      "epoch": 7.72106171516991,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 32434
    },
    {
      "epoch": 7.721299767898589,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 32435
    },
    {
      "epoch": 7.721537820627269,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 32436
    },
    {
      "epoch": 7.7217758733559485,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.284,
      "step": 32437
    },
    {
      "epoch": 7.722013926084628,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 32438
    },
    {
      "epoch": 7.722251978813307,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0503,
      "step": 32439
    },
    {
      "epoch": 7.722490031541986,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 32440
    },
    {
      "epoch": 7.722728084270666,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 32441
    },
    {
      "epoch": 7.722966136999346,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 32442
    },
    {
      "epoch": 7.723204189728024,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 32443
    },
    {
      "epoch": 7.723442242456704,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 32444
    },
    {
      "epoch": 7.723680295185384,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 32445
    },
    {
      "epoch": 7.723918347914063,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 32446
    },
    {
      "epoch": 7.724156400642743,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 32447
    },
    {
      "epoch": 7.7243944533714215,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 32448
    },
    {
      "epoch": 7.724632506100101,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3704,
      "step": 32449
    },
    {
      "epoch": 7.724870558828781,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3048,
      "step": 32450
    },
    {
      "epoch": 7.72510861155746,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3156,
      "step": 32451
    },
    {
      "epoch": 7.725346664286139,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 32452
    },
    {
      "epoch": 7.725584717014819,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 32453
    },
    {
      "epoch": 7.725822769743498,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 32454
    },
    {
      "epoch": 7.726060822472178,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9569,
      "step": 32455
    },
    {
      "epoch": 7.7262988752008575,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 32456
    },
    {
      "epoch": 7.726536927929536,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 32457
    },
    {
      "epoch": 7.726774980658216,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 32458
    },
    {
      "epoch": 7.727013033386895,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.5105,
      "step": 32459
    },
    {
      "epoch": 7.727251086115575,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9016,
      "step": 32460
    },
    {
      "epoch": 7.727489138844254,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 32461
    },
    {
      "epoch": 7.727727191572933,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 32462
    },
    {
      "epoch": 7.727965244301613,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 32463
    },
    {
      "epoch": 7.7282032970302925,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9,
      "step": 32464
    },
    {
      "epoch": 7.728441349758971,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2783,
      "step": 32465
    },
    {
      "epoch": 7.728679402487651,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 32466
    },
    {
      "epoch": 7.7289174552163304,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 32467
    },
    {
      "epoch": 7.72915550794501,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 32468
    },
    {
      "epoch": 7.729393560673689,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 32469
    },
    {
      "epoch": 7.729631613402368,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 32470
    },
    {
      "epoch": 7.729869666131048,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 32471
    },
    {
      "epoch": 7.730107718859728,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 32472
    },
    {
      "epoch": 7.730345771588407,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 32473
    },
    {
      "epoch": 7.730583824317086,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 32474
    },
    {
      "epoch": 7.7308218770457655,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 32475
    },
    {
      "epoch": 7.731059929774445,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 32476
    },
    {
      "epoch": 7.731297982503125,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 32477
    },
    {
      "epoch": 7.731536035231803,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 32478
    },
    {
      "epoch": 7.731774087960483,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 32479
    },
    {
      "epoch": 7.732012140689163,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 32480
    },
    {
      "epoch": 7.732250193417842,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 32481
    },
    {
      "epoch": 7.732488246146522,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 32482
    },
    {
      "epoch": 7.732726298875201,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 32483
    },
    {
      "epoch": 7.73296435160388,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 32484
    },
    {
      "epoch": 7.73320240433256,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 32485
    },
    {
      "epoch": 7.733440457061239,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 32486
    },
    {
      "epoch": 7.733678509789918,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 32487
    },
    {
      "epoch": 7.733916562518598,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 32488
    },
    {
      "epoch": 7.734154615247277,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 32489
    },
    {
      "epoch": 7.734392667975957,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 32490
    },
    {
      "epoch": 7.7346307207046365,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 32491
    },
    {
      "epoch": 7.734868773433315,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 32492
    },
    {
      "epoch": 7.735106826161995,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2727,
      "step": 32493
    },
    {
      "epoch": 7.7353448788906745,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 32494
    },
    {
      "epoch": 7.735582931619354,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 32495
    },
    {
      "epoch": 7.735820984348033,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8376,
      "step": 32496
    },
    {
      "epoch": 7.736059037076712,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 32497
    },
    {
      "epoch": 7.736297089805392,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 32498
    },
    {
      "epoch": 7.736535142534072,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 32499
    },
    {
      "epoch": 7.73677319526275,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1275,
      "step": 32500
    },
    {
      "epoch": 7.73677319526275,
      "eval_runtime": 0.5369,
      "eval_samples_per_second": 1.863,
      "eval_steps_per_second": 1.863,
      "step": 32500
    },
    {
      "epoch": 7.73701124799143,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 32501
    },
    {
      "epoch": 7.7372493007201095,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 32502
    },
    {
      "epoch": 7.737487353448789,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.031,
      "step": 32503
    },
    {
      "epoch": 7.737725406177468,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 32504
    },
    {
      "epoch": 7.7379634589061475,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 32505
    },
    {
      "epoch": 7.738201511634827,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 32506
    },
    {
      "epoch": 7.738439564363507,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 32507
    },
    {
      "epoch": 7.738677617092186,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 32508
    },
    {
      "epoch": 7.738915669820865,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 32509
    },
    {
      "epoch": 7.739153722549545,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7065,
      "step": 32510
    },
    {
      "epoch": 7.739391775278224,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 32511
    },
    {
      "epoch": 7.739629828006904,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 32512
    },
    {
      "epoch": 7.7398678807355825,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 32513
    },
    {
      "epoch": 7.740105933464262,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.5942,
      "step": 32514
    },
    {
      "epoch": 7.740343986192942,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 32515
    },
    {
      "epoch": 7.740582038921621,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 32516
    },
    {
      "epoch": 7.740820091650301,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 32517
    },
    {
      "epoch": 7.74105814437898,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 32518
    },
    {
      "epoch": 7.741296197107659,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 32519
    },
    {
      "epoch": 7.741534249836339,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 32520
    },
    {
      "epoch": 7.7417723025650185,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 32521
    },
    {
      "epoch": 7.742010355293697,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.767,
      "step": 32522
    },
    {
      "epoch": 7.742248408022377,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 32523
    },
    {
      "epoch": 7.742486460751056,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 32524
    },
    {
      "epoch": 7.742724513479736,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 32525
    },
    {
      "epoch": 7.742962566208416,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1713,
      "step": 32526
    },
    {
      "epoch": 7.743200618937094,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 32527
    },
    {
      "epoch": 7.743438671665774,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 32528
    },
    {
      "epoch": 7.7436767243944535,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 32529
    },
    {
      "epoch": 7.743914777123133,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 32530
    },
    {
      "epoch": 7.744152829851812,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 32531
    },
    {
      "epoch": 7.7443908825804915,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 32532
    },
    {
      "epoch": 7.744628935309171,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 32533
    },
    {
      "epoch": 7.744866988037851,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 32534
    },
    {
      "epoch": 7.74510504076653,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1927,
      "step": 32535
    },
    {
      "epoch": 7.745343093495209,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 32536
    },
    {
      "epoch": 7.745581146223889,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0828,
      "step": 32537
    },
    {
      "epoch": 7.745819198952568,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 32538
    },
    {
      "epoch": 7.746057251681247,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 32539
    },
    {
      "epoch": 7.7462953044099265,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 32540
    },
    {
      "epoch": 7.746533357138606,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 32541
    },
    {
      "epoch": 7.746771409867286,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 32542
    },
    {
      "epoch": 7.747009462595965,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 32543
    },
    {
      "epoch": 7.747247515324644,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 32544
    },
    {
      "epoch": 7.747485568053324,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 32545
    },
    {
      "epoch": 7.747723620782003,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 32546
    },
    {
      "epoch": 7.747961673510683,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 32547
    },
    {
      "epoch": 7.748199726239362,
      "grad_norm": 2.65625,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 32548
    },
    {
      "epoch": 7.748437778968041,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 32549
    },
    {
      "epoch": 7.748675831696721,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 32550
    },
    {
      "epoch": 7.7489138844254,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 32551
    },
    {
      "epoch": 7.74915193715408,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 32552
    },
    {
      "epoch": 7.749389989882759,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 32553
    },
    {
      "epoch": 7.749628042611438,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 32554
    },
    {
      "epoch": 7.749866095340118,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 32555
    },
    {
      "epoch": 7.7501041480687975,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 32556
    },
    {
      "epoch": 7.750342200797476,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1036,
      "step": 32557
    },
    {
      "epoch": 7.750580253526156,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 32558
    },
    {
      "epoch": 7.7508183062548355,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 32559
    },
    {
      "epoch": 7.751056358983515,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 32560
    },
    {
      "epoch": 7.751294411712195,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 32561
    },
    {
      "epoch": 7.751532464440873,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 32562
    },
    {
      "epoch": 7.751770517169553,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 32563
    },
    {
      "epoch": 7.752008569898233,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 32564
    },
    {
      "epoch": 7.752246622626912,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.394,
      "step": 32565
    },
    {
      "epoch": 7.752484675355591,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 32566
    },
    {
      "epoch": 7.7527227280842705,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 32567
    },
    {
      "epoch": 7.75296078081295,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 32568
    },
    {
      "epoch": 7.75319883354163,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 32569
    },
    {
      "epoch": 7.753436886270309,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 32570
    },
    {
      "epoch": 7.753674938998988,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 32571
    },
    {
      "epoch": 7.753912991727668,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 32572
    },
    {
      "epoch": 7.754151044456347,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7753,
      "step": 32573
    },
    {
      "epoch": 7.754389097185026,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 32574
    },
    {
      "epoch": 7.754627149913706,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 32575
    },
    {
      "epoch": 7.754865202642385,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2242,
      "step": 32576
    },
    {
      "epoch": 7.755103255371065,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0083,
      "step": 32577
    },
    {
      "epoch": 7.755341308099744,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 32578
    },
    {
      "epoch": 7.755579360828423,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0675,
      "step": 32579
    },
    {
      "epoch": 7.755817413557103,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1125,
      "step": 32580
    },
    {
      "epoch": 7.756055466285782,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 32581
    },
    {
      "epoch": 7.756293519014462,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 32582
    },
    {
      "epoch": 7.756531571743141,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4085,
      "step": 32583
    },
    {
      "epoch": 7.75676962447182,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 32584
    },
    {
      "epoch": 7.7570076772005,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 32585
    },
    {
      "epoch": 7.7572457299291795,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 32586
    },
    {
      "epoch": 7.757483782657859,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 32587
    },
    {
      "epoch": 7.757721835386538,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 32588
    },
    {
      "epoch": 7.757959888115217,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 32589
    },
    {
      "epoch": 7.758197940843897,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 32590
    },
    {
      "epoch": 7.758435993572577,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 32591
    },
    {
      "epoch": 7.758674046301255,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 32592
    },
    {
      "epoch": 7.758912099029935,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 32593
    },
    {
      "epoch": 7.7591501517586146,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 32594
    },
    {
      "epoch": 7.759388204487294,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 32595
    },
    {
      "epoch": 7.759626257215974,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 32596
    },
    {
      "epoch": 7.7598643099446525,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 32597
    },
    {
      "epoch": 7.760102362673332,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 32598
    },
    {
      "epoch": 7.760340415402012,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 32599
    },
    {
      "epoch": 7.760578468130691,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 32600
    },
    {
      "epoch": 7.760578468130691,
      "eval_runtime": 0.5151,
      "eval_samples_per_second": 1.941,
      "eval_steps_per_second": 1.941,
      "step": 32600
    },
    {
      "epoch": 7.76081652085937,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0994,
      "step": 32601
    },
    {
      "epoch": 7.76105457358805,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 32602
    },
    {
      "epoch": 7.761292626316729,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 32603
    },
    {
      "epoch": 7.761530679045409,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 32604
    },
    {
      "epoch": 7.761768731774088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1468,
      "step": 32605
    },
    {
      "epoch": 7.762006784502767,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 32606
    },
    {
      "epoch": 7.762244837231447,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 32607
    },
    {
      "epoch": 7.762482889960126,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 32608
    },
    {
      "epoch": 7.762720942688805,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1809,
      "step": 32609
    },
    {
      "epoch": 7.762958995417485,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1384,
      "step": 32610
    },
    {
      "epoch": 7.763197048146164,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9435,
      "step": 32611
    },
    {
      "epoch": 7.763435100874844,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 32612
    },
    {
      "epoch": 7.7636731536035235,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 32613
    },
    {
      "epoch": 7.763911206332202,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 32614
    },
    {
      "epoch": 7.764149259060882,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7799,
      "step": 32615
    },
    {
      "epoch": 7.764387311789561,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7598,
      "step": 32616
    },
    {
      "epoch": 7.764625364518241,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 32617
    },
    {
      "epoch": 7.76486341724692,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 32618
    },
    {
      "epoch": 7.765101469975599,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 32619
    },
    {
      "epoch": 7.765339522704279,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.253,
      "step": 32620
    },
    {
      "epoch": 7.765577575432959,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 32621
    },
    {
      "epoch": 7.765815628161638,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 32622
    },
    {
      "epoch": 7.766053680890317,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8701,
      "step": 32623
    },
    {
      "epoch": 7.7662917336189965,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.992,
      "step": 32624
    },
    {
      "epoch": 7.766529786347676,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 32625
    },
    {
      "epoch": 7.766767839076356,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 32626
    },
    {
      "epoch": 7.767005891805034,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 32627
    },
    {
      "epoch": 7.767243944533714,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 32628
    },
    {
      "epoch": 7.767481997262394,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1637,
      "step": 32629
    },
    {
      "epoch": 7.767720049991073,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 32630
    },
    {
      "epoch": 7.767958102719753,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 32631
    },
    {
      "epoch": 7.768196155448432,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8316,
      "step": 32632
    },
    {
      "epoch": 7.768434208177111,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2869,
      "step": 32633
    },
    {
      "epoch": 7.768672260905791,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 32634
    },
    {
      "epoch": 7.76891031363447,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.259,
      "step": 32635
    },
    {
      "epoch": 7.769148366363149,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 32636
    },
    {
      "epoch": 7.769386419091829,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 32637
    },
    {
      "epoch": 7.769624471820508,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 32638
    },
    {
      "epoch": 7.769862524549188,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 32639
    },
    {
      "epoch": 7.7701005772778675,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 32640
    },
    {
      "epoch": 7.770338630006546,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8141,
      "step": 32641
    },
    {
      "epoch": 7.770576682735226,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 32642
    },
    {
      "epoch": 7.770814735463905,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 32643
    },
    {
      "epoch": 7.771052788192584,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 32644
    },
    {
      "epoch": 7.771290840921264,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 32645
    },
    {
      "epoch": 7.771528893649943,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 32646
    },
    {
      "epoch": 7.771766946378623,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 32647
    },
    {
      "epoch": 7.772004999107303,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1177,
      "step": 32648
    },
    {
      "epoch": 7.772243051835981,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 32649
    },
    {
      "epoch": 7.772481104564661,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 32650
    },
    {
      "epoch": 7.7727191572933405,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 32651
    },
    {
      "epoch": 7.77295721002202,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 32652
    },
    {
      "epoch": 7.773195262750699,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1737,
      "step": 32653
    },
    {
      "epoch": 7.773433315479378,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4722,
      "step": 32654
    },
    {
      "epoch": 7.773671368208058,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 32655
    },
    {
      "epoch": 7.773909420936738,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 32656
    },
    {
      "epoch": 7.774147473665417,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 32657
    },
    {
      "epoch": 7.774385526394096,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 32658
    },
    {
      "epoch": 7.774623579122776,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 32659
    },
    {
      "epoch": 7.774861631851455,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0847,
      "step": 32660
    },
    {
      "epoch": 7.775099684580135,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 32661
    },
    {
      "epoch": 7.7753377373088135,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 32662
    },
    {
      "epoch": 7.775575790037493,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2577,
      "step": 32663
    },
    {
      "epoch": 7.775813842766173,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 32664
    },
    {
      "epoch": 7.776051895494852,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7693,
      "step": 32665
    },
    {
      "epoch": 7.776289948223532,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 32666
    },
    {
      "epoch": 7.776528000952211,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 32667
    },
    {
      "epoch": 7.77676605368089,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 32668
    },
    {
      "epoch": 7.77700410640957,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.431,
      "step": 32669
    },
    {
      "epoch": 7.7772421591382495,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 32670
    },
    {
      "epoch": 7.777480211866928,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 32671
    },
    {
      "epoch": 7.777718264595608,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 32672
    },
    {
      "epoch": 7.777956317324287,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 32673
    },
    {
      "epoch": 7.778194370052967,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8793,
      "step": 32674
    },
    {
      "epoch": 7.778432422781647,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 32675
    },
    {
      "epoch": 7.778670475510325,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 32676
    },
    {
      "epoch": 7.778908528239005,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 32677
    },
    {
      "epoch": 7.7791465809676845,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 32678
    },
    {
      "epoch": 7.779384633696364,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 32679
    },
    {
      "epoch": 7.779622686425043,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3138,
      "step": 32680
    },
    {
      "epoch": 7.7798607391537224,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 32681
    },
    {
      "epoch": 7.780098791882402,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 32682
    },
    {
      "epoch": 7.780336844611082,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 32683
    },
    {
      "epoch": 7.78057489733976,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 32684
    },
    {
      "epoch": 7.78081295006844,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 32685
    },
    {
      "epoch": 7.78105100279712,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 32686
    },
    {
      "epoch": 7.781289055525799,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 32687
    },
    {
      "epoch": 7.781527108254478,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 32688
    },
    {
      "epoch": 7.7817651609831575,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.549,
      "step": 32689
    },
    {
      "epoch": 7.782003213711837,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1005,
      "step": 32690
    },
    {
      "epoch": 7.782241266440517,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 32691
    },
    {
      "epoch": 7.782479319169196,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 32692
    },
    {
      "epoch": 7.782717371897875,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1716,
      "step": 32693
    },
    {
      "epoch": 7.782955424626555,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 32694
    },
    {
      "epoch": 7.783193477355234,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 32695
    },
    {
      "epoch": 7.783431530083914,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.5387,
      "step": 32696
    },
    {
      "epoch": 7.783669582812593,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 32697
    },
    {
      "epoch": 7.783907635541272,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 32698
    },
    {
      "epoch": 7.784145688269952,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 32699
    },
    {
      "epoch": 7.784383740998631,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 32700
    },
    {
      "epoch": 7.784383740998631,
      "eval_runtime": 0.546,
      "eval_samples_per_second": 1.831,
      "eval_steps_per_second": 1.831,
      "step": 32700
    },
    {
      "epoch": 7.784621793727311,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 32701
    },
    {
      "epoch": 7.78485984645599,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 32702
    },
    {
      "epoch": 7.785097899184669,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1218,
      "step": 32703
    },
    {
      "epoch": 7.785335951913349,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 32704
    },
    {
      "epoch": 7.7855740046420285,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 32705
    },
    {
      "epoch": 7.785812057370707,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.011,
      "step": 32706
    },
    {
      "epoch": 7.786050110099387,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 32707
    },
    {
      "epoch": 7.7862881628280665,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 32708
    },
    {
      "epoch": 7.786526215556746,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 32709
    },
    {
      "epoch": 7.786764268285426,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 32710
    },
    {
      "epoch": 7.787002321014104,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 32711
    },
    {
      "epoch": 7.787240373742784,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 32712
    },
    {
      "epoch": 7.787478426471464,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 32713
    },
    {
      "epoch": 7.787716479200143,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 32714
    },
    {
      "epoch": 7.787954531928822,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 32715
    },
    {
      "epoch": 7.7881925846575015,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 32716
    },
    {
      "epoch": 7.788430637386181,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 32717
    },
    {
      "epoch": 7.788668690114861,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 32718
    },
    {
      "epoch": 7.78890674284354,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2801,
      "step": 32719
    },
    {
      "epoch": 7.789144795572219,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 32720
    },
    {
      "epoch": 7.789382848300899,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9321,
      "step": 32721
    },
    {
      "epoch": 7.789620901029578,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 32722
    },
    {
      "epoch": 7.789858953758257,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 32723
    },
    {
      "epoch": 7.790097006486937,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 32724
    },
    {
      "epoch": 7.790335059215616,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 32725
    },
    {
      "epoch": 7.790573111944296,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 32726
    },
    {
      "epoch": 7.790811164672975,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2217,
      "step": 32727
    },
    {
      "epoch": 7.791049217401654,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3459,
      "step": 32728
    },
    {
      "epoch": 7.791287270130334,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8214,
      "step": 32729
    },
    {
      "epoch": 7.791525322859013,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 32730
    },
    {
      "epoch": 7.791763375587693,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7874,
      "step": 32731
    },
    {
      "epoch": 7.792001428316372,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 32732
    },
    {
      "epoch": 7.792239481045051,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4249,
      "step": 32733
    },
    {
      "epoch": 7.792477533773731,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 32734
    },
    {
      "epoch": 7.7927155865024105,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 32735
    },
    {
      "epoch": 7.79295363923109,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 32736
    },
    {
      "epoch": 7.793191691959769,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 32737
    },
    {
      "epoch": 7.793429744688448,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 32738
    },
    {
      "epoch": 7.793667797417128,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 32739
    },
    {
      "epoch": 7.793905850145808,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 32740
    },
    {
      "epoch": 7.794143902874486,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 32741
    },
    {
      "epoch": 7.794381955603166,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 32742
    },
    {
      "epoch": 7.7946200083318455,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 32743
    },
    {
      "epoch": 7.794858061060525,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 32744
    },
    {
      "epoch": 7.795096113789205,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 32745
    },
    {
      "epoch": 7.7953341665178835,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 32746
    },
    {
      "epoch": 7.795572219246563,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2374,
      "step": 32747
    },
    {
      "epoch": 7.795810271975243,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2094,
      "step": 32748
    },
    {
      "epoch": 7.796048324703922,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 32749
    },
    {
      "epoch": 7.796286377432601,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 32750
    },
    {
      "epoch": 7.796524430161281,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 32751
    },
    {
      "epoch": 7.79676248288996,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 32752
    },
    {
      "epoch": 7.79700053561864,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7647,
      "step": 32753
    },
    {
      "epoch": 7.797238588347319,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 32754
    },
    {
      "epoch": 7.797476641075998,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6724,
      "step": 32755
    },
    {
      "epoch": 7.797714693804678,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 32756
    },
    {
      "epoch": 7.797952746533357,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 32757
    },
    {
      "epoch": 7.798190799262036,
      "grad_norm": 2.90625,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 32758
    },
    {
      "epoch": 7.798428851990716,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9669,
      "step": 32759
    },
    {
      "epoch": 7.798666904719395,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 32760
    },
    {
      "epoch": 7.798904957448075,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4001,
      "step": 32761
    },
    {
      "epoch": 7.7991430101767545,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 32762
    },
    {
      "epoch": 7.799381062905433,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 32763
    },
    {
      "epoch": 7.799619115634113,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 32764
    },
    {
      "epoch": 7.799857168362792,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 32765
    },
    {
      "epoch": 7.800095221091472,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3037,
      "step": 32766
    },
    {
      "epoch": 7.800333273820151,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 32767
    },
    {
      "epoch": 7.80057132654883,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 32768
    },
    {
      "epoch": 7.80080937927751,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 32769
    },
    {
      "epoch": 7.8010474320061896,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7226,
      "step": 32770
    },
    {
      "epoch": 7.801285484734869,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 32771
    },
    {
      "epoch": 7.801523537463548,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2294,
      "step": 32772
    },
    {
      "epoch": 7.8017615901922275,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2979,
      "step": 32773
    },
    {
      "epoch": 7.801999642920907,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 32774
    },
    {
      "epoch": 7.802237695649587,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 32775
    },
    {
      "epoch": 7.802475748378265,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8037,
      "step": 32776
    },
    {
      "epoch": 7.802713801106945,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 32777
    },
    {
      "epoch": 7.802951853835625,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9244,
      "step": 32778
    },
    {
      "epoch": 7.803189906564304,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 32779
    },
    {
      "epoch": 7.803427959292984,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 32780
    },
    {
      "epoch": 7.8036660120216625,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4923,
      "step": 32781
    },
    {
      "epoch": 7.803904064750342,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 32782
    },
    {
      "epoch": 7.804142117479022,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 32783
    },
    {
      "epoch": 7.804380170207701,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1963,
      "step": 32784
    },
    {
      "epoch": 7.80461822293638,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 1.9883,
      "step": 32785
    },
    {
      "epoch": 7.80485627566506,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3678,
      "step": 32786
    },
    {
      "epoch": 7.805094328393739,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 32787
    },
    {
      "epoch": 7.805332381122419,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3699,
      "step": 32788
    },
    {
      "epoch": 7.8055704338510985,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 32789
    },
    {
      "epoch": 7.805808486579777,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0505,
      "step": 32790
    },
    {
      "epoch": 7.806046539308457,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2472,
      "step": 32791
    },
    {
      "epoch": 7.806284592037136,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 32792
    },
    {
      "epoch": 7.806522644765815,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 32793
    },
    {
      "epoch": 7.806760697494495,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 32794
    },
    {
      "epoch": 7.806998750223174,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 32795
    },
    {
      "epoch": 7.807236802951854,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 32796
    },
    {
      "epoch": 7.807474855680534,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 32797
    },
    {
      "epoch": 7.807712908409212,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2114,
      "step": 32798
    },
    {
      "epoch": 7.807950961137892,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 32799
    },
    {
      "epoch": 7.8081890138665715,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.956,
      "step": 32800
    },
    {
      "epoch": 7.8081890138665715,
      "eval_runtime": 0.5276,
      "eval_samples_per_second": 1.895,
      "eval_steps_per_second": 1.895,
      "step": 32800
    },
    {
      "epoch": 7.808427066595251,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 32801
    },
    {
      "epoch": 7.80866511932393,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 32802
    },
    {
      "epoch": 7.808903172052609,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 32803
    },
    {
      "epoch": 7.809141224781289,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 32804
    },
    {
      "epoch": 7.809379277509969,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2418,
      "step": 32805
    },
    {
      "epoch": 7.809617330238648,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 32806
    },
    {
      "epoch": 7.809855382967327,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 32807
    },
    {
      "epoch": 7.810093435696007,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 32808
    },
    {
      "epoch": 7.810331488424686,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2488,
      "step": 32809
    },
    {
      "epoch": 7.810569541153366,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 32810
    },
    {
      "epoch": 7.8108075938820445,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3196,
      "step": 32811
    },
    {
      "epoch": 7.811045646610724,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 32812
    },
    {
      "epoch": 7.811283699339404,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 32813
    },
    {
      "epoch": 7.811521752068083,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 32814
    },
    {
      "epoch": 7.811759804796763,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 32815
    },
    {
      "epoch": 7.811997857525442,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 32816
    },
    {
      "epoch": 7.812235910254121,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 32817
    },
    {
      "epoch": 7.812473962982801,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 32818
    },
    {
      "epoch": 7.81271201571148,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 32819
    },
    {
      "epoch": 7.812950068440159,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 32820
    },
    {
      "epoch": 7.813188121168839,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 32821
    },
    {
      "epoch": 7.813426173897518,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 32822
    },
    {
      "epoch": 7.813664226626198,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4336,
      "step": 32823
    },
    {
      "epoch": 7.813902279354878,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 32824
    },
    {
      "epoch": 7.814140332083556,
      "grad_norm": 2.765625,
      "learning_rate": 3e-05,
      "loss": 2.0804,
      "step": 32825
    },
    {
      "epoch": 7.814378384812236,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 32826
    },
    {
      "epoch": 7.8146164375409155,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 32827
    },
    {
      "epoch": 7.814854490269594,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 32828
    },
    {
      "epoch": 7.815092542998274,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 32829
    },
    {
      "epoch": 7.815330595726953,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 32830
    },
    {
      "epoch": 7.815568648455633,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 32831
    },
    {
      "epoch": 7.815806701184313,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 32832
    },
    {
      "epoch": 7.816044753912991,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 32833
    },
    {
      "epoch": 7.816282806641671,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 32834
    },
    {
      "epoch": 7.816520859370351,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 32835
    },
    {
      "epoch": 7.81675891209903,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 32836
    },
    {
      "epoch": 7.816996964827709,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 32837
    },
    {
      "epoch": 7.8172350175563885,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 32838
    },
    {
      "epoch": 7.817473070285068,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 32839
    },
    {
      "epoch": 7.817711123013748,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.5487,
      "step": 32840
    },
    {
      "epoch": 7.817949175742427,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 32841
    },
    {
      "epoch": 7.818187228471106,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7675,
      "step": 32842
    },
    {
      "epoch": 7.818425281199786,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 32843
    },
    {
      "epoch": 7.818663333928465,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3351,
      "step": 32844
    },
    {
      "epoch": 7.818901386657145,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0177,
      "step": 32845
    },
    {
      "epoch": 7.819139439385824,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 32846
    },
    {
      "epoch": 7.819377492114503,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 32847
    },
    {
      "epoch": 7.819615544843183,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 32848
    },
    {
      "epoch": 7.819853597571862,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 32849
    },
    {
      "epoch": 7.820091650300542,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 32850
    },
    {
      "epoch": 7.820329703029221,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 32851
    },
    {
      "epoch": 7.8205677557579,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0127,
      "step": 32852
    },
    {
      "epoch": 7.82080580848658,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 32853
    },
    {
      "epoch": 7.8210438612152595,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1602,
      "step": 32854
    },
    {
      "epoch": 7.821281913943938,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1485,
      "step": 32855
    },
    {
      "epoch": 7.821519966672618,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 32856
    },
    {
      "epoch": 7.8217580194012974,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 32857
    },
    {
      "epoch": 7.821996072129977,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 32858
    },
    {
      "epoch": 7.822234124858657,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 32859
    },
    {
      "epoch": 7.822472177587335,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 32860
    },
    {
      "epoch": 7.822710230316015,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 32861
    },
    {
      "epoch": 7.822948283044695,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 32862
    },
    {
      "epoch": 7.823186335773374,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 32863
    },
    {
      "epoch": 7.823424388502053,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2771,
      "step": 32864
    },
    {
      "epoch": 7.8236624412307325,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 32865
    },
    {
      "epoch": 7.823900493959412,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 32866
    },
    {
      "epoch": 7.824138546688092,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 32867
    },
    {
      "epoch": 7.82437659941677,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 32868
    },
    {
      "epoch": 7.82461465214545,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1251,
      "step": 32869
    },
    {
      "epoch": 7.82485270487413,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 32870
    },
    {
      "epoch": 7.825090757602809,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 32871
    },
    {
      "epoch": 7.825328810331488,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 32872
    },
    {
      "epoch": 7.825566863060168,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 32873
    },
    {
      "epoch": 7.825804915788847,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 32874
    },
    {
      "epoch": 7.826042968517527,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3334,
      "step": 32875
    },
    {
      "epoch": 7.826281021246206,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6968,
      "step": 32876
    },
    {
      "epoch": 7.826519073974885,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9534,
      "step": 32877
    },
    {
      "epoch": 7.826757126703565,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 32878
    },
    {
      "epoch": 7.826995179432244,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0517,
      "step": 32879
    },
    {
      "epoch": 7.827233232160924,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 32880
    },
    {
      "epoch": 7.827471284889603,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 32881
    },
    {
      "epoch": 7.827709337618282,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1965,
      "step": 32882
    },
    {
      "epoch": 7.827947390346962,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2015,
      "step": 32883
    },
    {
      "epoch": 7.8281854430756415,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 32884
    },
    {
      "epoch": 7.828423495804321,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 32885
    },
    {
      "epoch": 7.828661548533,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 32886
    },
    {
      "epoch": 7.828899601261679,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 32887
    },
    {
      "epoch": 7.829137653990359,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 32888
    },
    {
      "epoch": 7.829375706719039,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 32889
    },
    {
      "epoch": 7.829613759447717,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 32890
    },
    {
      "epoch": 7.829851812176397,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 32891
    },
    {
      "epoch": 7.8300898649050765,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 32892
    },
    {
      "epoch": 7.830327917633756,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 32893
    },
    {
      "epoch": 7.830565970362436,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 32894
    },
    {
      "epoch": 7.8308040230911145,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 32895
    },
    {
      "epoch": 7.831042075819794,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 32896
    },
    {
      "epoch": 7.831280128548474,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 32897
    },
    {
      "epoch": 7.831518181277153,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 32898
    },
    {
      "epoch": 7.831756234005832,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 32899
    },
    {
      "epoch": 7.831994286734512,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 32900
    },
    {
      "epoch": 7.831994286734512,
      "eval_runtime": 0.5294,
      "eval_samples_per_second": 1.889,
      "eval_steps_per_second": 1.889,
      "step": 32900
    },
    {
      "epoch": 7.832232339463191,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 32901
    },
    {
      "epoch": 7.832470392191871,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0665,
      "step": 32902
    },
    {
      "epoch": 7.8327084449205495,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1262,
      "step": 32903
    },
    {
      "epoch": 7.832946497649229,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9945,
      "step": 32904
    },
    {
      "epoch": 7.833184550377909,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 32905
    },
    {
      "epoch": 7.833422603106588,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 32906
    },
    {
      "epoch": 7.833660655835267,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1704,
      "step": 32907
    },
    {
      "epoch": 7.833898708563947,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 32908
    },
    {
      "epoch": 7.834136761292626,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 32909
    },
    {
      "epoch": 7.834374814021306,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 32910
    },
    {
      "epoch": 7.8346128667499855,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0678,
      "step": 32911
    },
    {
      "epoch": 7.834850919478664,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 32912
    },
    {
      "epoch": 7.835088972207344,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.4587,
      "step": 32913
    },
    {
      "epoch": 7.835327024936023,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9194,
      "step": 32914
    },
    {
      "epoch": 7.835565077664703,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0178,
      "step": 32915
    },
    {
      "epoch": 7.835803130393382,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 32916
    },
    {
      "epoch": 7.836041183122061,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 32917
    },
    {
      "epoch": 7.836279235850741,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.899,
      "step": 32918
    },
    {
      "epoch": 7.8365172885794205,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 32919
    },
    {
      "epoch": 7.8367553413081,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 32920
    },
    {
      "epoch": 7.836993394036779,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9665,
      "step": 32921
    },
    {
      "epoch": 7.8372314467654585,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1853,
      "step": 32922
    },
    {
      "epoch": 7.837469499494138,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3166,
      "step": 32923
    },
    {
      "epoch": 7.837707552222818,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2968,
      "step": 32924
    },
    {
      "epoch": 7.837945604951496,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2574,
      "step": 32925
    },
    {
      "epoch": 7.838183657680176,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 32926
    },
    {
      "epoch": 7.838421710408856,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 32927
    },
    {
      "epoch": 7.838659763137535,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 32928
    },
    {
      "epoch": 7.838897815866215,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 32929
    },
    {
      "epoch": 7.8391358685948935,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 32930
    },
    {
      "epoch": 7.839373921323573,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 32931
    },
    {
      "epoch": 7.839611974052253,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 32932
    },
    {
      "epoch": 7.839850026780932,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 32933
    },
    {
      "epoch": 7.840088079509611,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 32934
    },
    {
      "epoch": 7.840326132238291,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 32935
    },
    {
      "epoch": 7.84056418496697,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3654,
      "step": 32936
    },
    {
      "epoch": 7.84080223769565,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 32937
    },
    {
      "epoch": 7.8410402904243295,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 32938
    },
    {
      "epoch": 7.841278343153008,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 32939
    },
    {
      "epoch": 7.841516395881688,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 32940
    },
    {
      "epoch": 7.841754448610367,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2368,
      "step": 32941
    },
    {
      "epoch": 7.841992501339046,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 32942
    },
    {
      "epoch": 7.842230554067726,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 32943
    },
    {
      "epoch": 7.842468606796405,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 32944
    },
    {
      "epoch": 7.842706659525085,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 32945
    },
    {
      "epoch": 7.8429447122537645,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1192,
      "step": 32946
    },
    {
      "epoch": 7.843182764982443,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1749,
      "step": 32947
    },
    {
      "epoch": 7.843420817711123,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 32948
    },
    {
      "epoch": 7.8436588704398025,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 32949
    },
    {
      "epoch": 7.843896923168482,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 32950
    },
    {
      "epoch": 7.844134975897161,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 32951
    },
    {
      "epoch": 7.84437302862584,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 32952
    },
    {
      "epoch": 7.84461108135452,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4962,
      "step": 32953
    },
    {
      "epoch": 7.8448491340832,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 32954
    },
    {
      "epoch": 7.845087186811879,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 32955
    },
    {
      "epoch": 7.845325239540558,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.364,
      "step": 32956
    },
    {
      "epoch": 7.8455632922692375,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2713,
      "step": 32957
    },
    {
      "epoch": 7.845801344997917,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 32958
    },
    {
      "epoch": 7.846039397726597,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3582,
      "step": 32959
    },
    {
      "epoch": 7.8462774504552755,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 32960
    },
    {
      "epoch": 7.846515503183955,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 32961
    },
    {
      "epoch": 7.846753555912635,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 32962
    },
    {
      "epoch": 7.846991608641314,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 32963
    },
    {
      "epoch": 7.847229661369994,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7306,
      "step": 32964
    },
    {
      "epoch": 7.847467714098673,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 32965
    },
    {
      "epoch": 7.847705766827352,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.235,
      "step": 32966
    },
    {
      "epoch": 7.847943819556032,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 32967
    },
    {
      "epoch": 7.848181872284711,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8947,
      "step": 32968
    },
    {
      "epoch": 7.84841992501339,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 32969
    },
    {
      "epoch": 7.84865797774207,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 32970
    },
    {
      "epoch": 7.848896030470749,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 32971
    },
    {
      "epoch": 7.849134083199429,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 32972
    },
    {
      "epoch": 7.849372135928109,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 32973
    },
    {
      "epoch": 7.849610188656787,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 32974
    },
    {
      "epoch": 7.849848241385467,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 32975
    },
    {
      "epoch": 7.8500862941141465,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7434,
      "step": 32976
    },
    {
      "epoch": 7.850324346842825,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 32977
    },
    {
      "epoch": 7.850562399571505,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0358,
      "step": 32978
    },
    {
      "epoch": 7.850800452300184,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4456,
      "step": 32979
    },
    {
      "epoch": 7.851038505028864,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 32980
    },
    {
      "epoch": 7.851276557757544,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 32981
    },
    {
      "epoch": 7.851514610486222,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2834,
      "step": 32982
    },
    {
      "epoch": 7.851752663214902,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3413,
      "step": 32983
    },
    {
      "epoch": 7.8519907159435816,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 32984
    },
    {
      "epoch": 7.852228768672261,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1975,
      "step": 32985
    },
    {
      "epoch": 7.85246682140094,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 32986
    },
    {
      "epoch": 7.8527048741296195,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1676,
      "step": 32987
    },
    {
      "epoch": 7.852942926858299,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 32988
    },
    {
      "epoch": 7.853180979586979,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 32989
    },
    {
      "epoch": 7.853419032315658,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 32990
    },
    {
      "epoch": 7.853657085044337,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7243,
      "step": 32991
    },
    {
      "epoch": 7.853895137773017,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 32992
    },
    {
      "epoch": 7.854133190501696,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 32993
    },
    {
      "epoch": 7.854371243230376,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 32994
    },
    {
      "epoch": 7.8546092959590545,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 32995
    },
    {
      "epoch": 7.854847348687734,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9419,
      "step": 32996
    },
    {
      "epoch": 7.855085401416414,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3018,
      "step": 32997
    },
    {
      "epoch": 7.855323454145093,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 32998
    },
    {
      "epoch": 7.855561506873773,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 32999
    },
    {
      "epoch": 7.855799559602452,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 33000
    },
    {
      "epoch": 7.855799559602452,
      "eval_runtime": 0.522,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 33000
    },
    {
      "epoch": 7.856037612331131,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 33001
    },
    {
      "epoch": 7.856275665059811,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 33002
    },
    {
      "epoch": 7.8565137177884905,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 33003
    },
    {
      "epoch": 7.856751770517169,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 33004
    },
    {
      "epoch": 7.856989823245849,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 33005
    },
    {
      "epoch": 7.857227875974528,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 33006
    },
    {
      "epoch": 7.857465928703208,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.5141,
      "step": 33007
    },
    {
      "epoch": 7.857703981431888,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 33008
    },
    {
      "epoch": 7.857942034160566,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 2.1367,
      "step": 33009
    },
    {
      "epoch": 7.858180086889246,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 33010
    },
    {
      "epoch": 7.858418139617926,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 33011
    },
    {
      "epoch": 7.858656192346604,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 33012
    },
    {
      "epoch": 7.858894245075284,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9096,
      "step": 33013
    },
    {
      "epoch": 7.8591322978039635,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 33014
    },
    {
      "epoch": 7.859370350532643,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 33015
    },
    {
      "epoch": 7.859608403261323,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 33016
    },
    {
      "epoch": 7.859846455990001,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 33017
    },
    {
      "epoch": 7.860084508718681,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9605,
      "step": 33018
    },
    {
      "epoch": 7.860322561447361,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 33019
    },
    {
      "epoch": 7.86056061417604,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 33020
    },
    {
      "epoch": 7.860798666904719,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3797,
      "step": 33021
    },
    {
      "epoch": 7.861036719633399,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7522,
      "step": 33022
    },
    {
      "epoch": 7.861274772362078,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 33023
    },
    {
      "epoch": 7.861512825090758,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 33024
    },
    {
      "epoch": 7.861750877819437,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 33025
    },
    {
      "epoch": 7.861988930548116,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 33026
    },
    {
      "epoch": 7.862226983276796,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7876,
      "step": 33027
    },
    {
      "epoch": 7.862465036005475,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9522,
      "step": 33028
    },
    {
      "epoch": 7.862703088734155,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.3554,
      "step": 33029
    },
    {
      "epoch": 7.862941141462834,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3228,
      "step": 33030
    },
    {
      "epoch": 7.863179194191513,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 33031
    },
    {
      "epoch": 7.863417246920193,
      "grad_norm": 6.40625,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 33032
    },
    {
      "epoch": 7.863655299648872,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 33033
    },
    {
      "epoch": 7.863893352377552,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7607,
      "step": 33034
    },
    {
      "epoch": 7.864131405106231,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 33035
    },
    {
      "epoch": 7.86436945783491,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2762,
      "step": 33036
    },
    {
      "epoch": 7.86460751056359,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 33037
    },
    {
      "epoch": 7.86484556329227,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.399,
      "step": 33038
    },
    {
      "epoch": 7.865083616020948,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 33039
    },
    {
      "epoch": 7.865321668749628,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 33040
    },
    {
      "epoch": 7.8655597214783075,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8713,
      "step": 33041
    },
    {
      "epoch": 7.865797774206987,
      "grad_norm": 6.75,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 33042
    },
    {
      "epoch": 7.866035826935667,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8556,
      "step": 33043
    },
    {
      "epoch": 7.866273879664345,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 33044
    },
    {
      "epoch": 7.866511932393025,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 33045
    },
    {
      "epoch": 7.866749985121705,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 33046
    },
    {
      "epoch": 7.866988037850383,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 33047
    },
    {
      "epoch": 7.867226090579063,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 33048
    },
    {
      "epoch": 7.867464143307743,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 33049
    },
    {
      "epoch": 7.867702196036422,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 33050
    },
    {
      "epoch": 7.867940248765102,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 33051
    },
    {
      "epoch": 7.8681783014937805,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 33052
    },
    {
      "epoch": 7.86841635422246,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 33053
    },
    {
      "epoch": 7.86865440695114,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 33054
    },
    {
      "epoch": 7.868892459679819,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2434,
      "step": 33055
    },
    {
      "epoch": 7.869130512408498,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8185,
      "step": 33056
    },
    {
      "epoch": 7.869368565137178,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8658,
      "step": 33057
    },
    {
      "epoch": 7.869606617865857,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 33058
    },
    {
      "epoch": 7.869844670594537,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5619,
      "step": 33059
    },
    {
      "epoch": 7.8700827233232165,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 33060
    },
    {
      "epoch": 7.870320776051895,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 33061
    },
    {
      "epoch": 7.870558828780575,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 33062
    },
    {
      "epoch": 7.870796881509254,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 33063
    },
    {
      "epoch": 7.871034934237934,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 33064
    },
    {
      "epoch": 7.871272986966613,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 33065
    },
    {
      "epoch": 7.871511039695292,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9094,
      "step": 33066
    },
    {
      "epoch": 7.871749092423972,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 33067
    },
    {
      "epoch": 7.8719871451526515,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7149,
      "step": 33068
    },
    {
      "epoch": 7.872225197881331,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 33069
    },
    {
      "epoch": 7.87246325061001,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 33070
    },
    {
      "epoch": 7.8727013033386894,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 33071
    },
    {
      "epoch": 7.872939356067369,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 33072
    },
    {
      "epoch": 7.873177408796049,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4252,
      "step": 33073
    },
    {
      "epoch": 7.873415461524727,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 33074
    },
    {
      "epoch": 7.873653514253407,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8612,
      "step": 33075
    },
    {
      "epoch": 7.873891566982087,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 33076
    },
    {
      "epoch": 7.874129619710766,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 33077
    },
    {
      "epoch": 7.874367672439446,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 33078
    },
    {
      "epoch": 7.8746057251681245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 33079
    },
    {
      "epoch": 7.874843777896804,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 33080
    },
    {
      "epoch": 7.875081830625484,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.4439,
      "step": 33081
    },
    {
      "epoch": 7.875319883354163,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 33082
    },
    {
      "epoch": 7.875557936082842,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3259,
      "step": 33083
    },
    {
      "epoch": 7.875795988811522,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 33084
    },
    {
      "epoch": 7.876034041540201,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4613,
      "step": 33085
    },
    {
      "epoch": 7.876272094268881,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0448,
      "step": 33086
    },
    {
      "epoch": 7.87651014699756,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 33087
    },
    {
      "epoch": 7.876748199726239,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 33088
    },
    {
      "epoch": 7.876986252454919,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 33089
    },
    {
      "epoch": 7.877224305183598,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0293,
      "step": 33090
    },
    {
      "epoch": 7.877462357912277,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7193,
      "step": 33091
    },
    {
      "epoch": 7.877700410640957,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 33092
    },
    {
      "epoch": 7.877938463369636,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 33093
    },
    {
      "epoch": 7.878176516098316,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.5636,
      "step": 33094
    },
    {
      "epoch": 7.8784145688269955,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0195,
      "step": 33095
    },
    {
      "epoch": 7.878652621555674,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 33096
    },
    {
      "epoch": 7.878890674284354,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0883,
      "step": 33097
    },
    {
      "epoch": 7.8791287270130335,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 33098
    },
    {
      "epoch": 7.879366779741713,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2021,
      "step": 33099
    },
    {
      "epoch": 7.879604832470392,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 33100
    },
    {
      "epoch": 7.879604832470392,
      "eval_runtime": 0.5269,
      "eval_samples_per_second": 1.898,
      "eval_steps_per_second": 1.898,
      "step": 33100
    },
    {
      "epoch": 7.879842885199071,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 33101
    },
    {
      "epoch": 7.880080937927751,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 33102
    },
    {
      "epoch": 7.880318990656431,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 33103
    },
    {
      "epoch": 7.88055704338511,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2048,
      "step": 33104
    },
    {
      "epoch": 7.880795096113789,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6376,
      "step": 33105
    },
    {
      "epoch": 7.8810331488424685,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8218,
      "step": 33106
    },
    {
      "epoch": 7.881271201571148,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 33107
    },
    {
      "epoch": 7.881509254299828,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2225,
      "step": 33108
    },
    {
      "epoch": 7.8817473070285065,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 33109
    },
    {
      "epoch": 7.881985359757186,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2471,
      "step": 33110
    },
    {
      "epoch": 7.882223412485866,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 33111
    },
    {
      "epoch": 7.882461465214545,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2625,
      "step": 33112
    },
    {
      "epoch": 7.882699517943225,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 33113
    },
    {
      "epoch": 7.882937570671904,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8213,
      "step": 33114
    },
    {
      "epoch": 7.883175623400583,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.143,
      "step": 33115
    },
    {
      "epoch": 7.883413676129263,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7628,
      "step": 33116
    },
    {
      "epoch": 7.883651728857942,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 33117
    },
    {
      "epoch": 7.883889781586621,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 33118
    },
    {
      "epoch": 7.884127834315301,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0048,
      "step": 33119
    },
    {
      "epoch": 7.88436588704398,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1213,
      "step": 33120
    },
    {
      "epoch": 7.88460393977266,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 33121
    },
    {
      "epoch": 7.884841992501339,
      "grad_norm": 7.0625,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 33122
    },
    {
      "epoch": 7.885080045230018,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 33123
    },
    {
      "epoch": 7.885318097958698,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 33124
    },
    {
      "epoch": 7.8855561506873775,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 33125
    },
    {
      "epoch": 7.885794203416056,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 33126
    },
    {
      "epoch": 7.886032256144736,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 33127
    },
    {
      "epoch": 7.886270308873415,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 33128
    },
    {
      "epoch": 7.886508361602095,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 33129
    },
    {
      "epoch": 7.886746414330775,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 33130
    },
    {
      "epoch": 7.886984467059453,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3496,
      "step": 33131
    },
    {
      "epoch": 7.887222519788133,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 33132
    },
    {
      "epoch": 7.8874605725168125,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2776,
      "step": 33133
    },
    {
      "epoch": 7.887698625245492,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.2906,
      "step": 33134
    },
    {
      "epoch": 7.887936677974171,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 33135
    },
    {
      "epoch": 7.8881747307028505,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8795,
      "step": 33136
    },
    {
      "epoch": 7.88841278343153,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 33137
    },
    {
      "epoch": 7.88865083616021,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 33138
    },
    {
      "epoch": 7.888888888888889,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 33139
    },
    {
      "epoch": 7.889126941617568,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 33140
    },
    {
      "epoch": 7.889364994346248,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 33141
    },
    {
      "epoch": 7.889603047074927,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 33142
    },
    {
      "epoch": 7.889841099803607,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 33143
    },
    {
      "epoch": 7.8900791525322855,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 33144
    },
    {
      "epoch": 7.890317205260965,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 33145
    },
    {
      "epoch": 7.890555257989645,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 33146
    },
    {
      "epoch": 7.890793310718324,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7424,
      "step": 33147
    },
    {
      "epoch": 7.891031363447004,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 33148
    },
    {
      "epoch": 7.891269416175683,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 33149
    },
    {
      "epoch": 7.891507468904362,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 33150
    },
    {
      "epoch": 7.891745521633042,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 33151
    },
    {
      "epoch": 7.8919835743617215,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 33152
    },
    {
      "epoch": 7.8922216270904,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.917,
      "step": 33153
    },
    {
      "epoch": 7.89245967981908,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 33154
    },
    {
      "epoch": 7.892697732547759,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.194,
      "step": 33155
    },
    {
      "epoch": 7.892935785276439,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 33156
    },
    {
      "epoch": 7.893173838005119,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3561,
      "step": 33157
    },
    {
      "epoch": 7.893411890733797,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 33158
    },
    {
      "epoch": 7.893649943462477,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 33159
    },
    {
      "epoch": 7.8938879961911566,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 33160
    },
    {
      "epoch": 7.894126048919835,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.736,
      "step": 33161
    },
    {
      "epoch": 7.894364101648515,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.241,
      "step": 33162
    },
    {
      "epoch": 7.8946021543771945,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 33163
    },
    {
      "epoch": 7.894840207105874,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 33164
    },
    {
      "epoch": 7.895078259834554,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7318,
      "step": 33165
    },
    {
      "epoch": 7.895316312563232,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 33166
    },
    {
      "epoch": 7.895554365291912,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 33167
    },
    {
      "epoch": 7.895792418020592,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 33168
    },
    {
      "epoch": 7.896030470749271,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 33169
    },
    {
      "epoch": 7.89626852347795,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 33170
    },
    {
      "epoch": 7.8965065762066295,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8659,
      "step": 33171
    },
    {
      "epoch": 7.896744628935309,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 33172
    },
    {
      "epoch": 7.896982681663989,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3897,
      "step": 33173
    },
    {
      "epoch": 7.897220734392668,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4766,
      "step": 33174
    },
    {
      "epoch": 7.897458787121347,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 33175
    },
    {
      "epoch": 7.897696839850027,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 33176
    },
    {
      "epoch": 7.897934892578706,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2208,
      "step": 33177
    },
    {
      "epoch": 7.898172945307386,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 33178
    },
    {
      "epoch": 7.898410998036065,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 33179
    },
    {
      "epoch": 7.898649050764744,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 33180
    },
    {
      "epoch": 7.898887103493424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 33181
    },
    {
      "epoch": 7.899125156222103,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2563,
      "step": 33182
    },
    {
      "epoch": 7.899363208950783,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8986,
      "step": 33183
    },
    {
      "epoch": 7.899601261679462,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 33184
    },
    {
      "epoch": 7.899839314408141,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 33185
    },
    {
      "epoch": 7.900077367136821,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 33186
    },
    {
      "epoch": 7.900315419865501,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 33187
    },
    {
      "epoch": 7.900553472594179,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3357,
      "step": 33188
    },
    {
      "epoch": 7.900791525322859,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9586,
      "step": 33189
    },
    {
      "epoch": 7.9010295780515385,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0337,
      "step": 33190
    },
    {
      "epoch": 7.901267630780218,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 33191
    },
    {
      "epoch": 7.901505683508898,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 33192
    },
    {
      "epoch": 7.901743736237576,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 33193
    },
    {
      "epoch": 7.901981788966256,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0276,
      "step": 33194
    },
    {
      "epoch": 7.902219841694936,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 33195
    },
    {
      "epoch": 7.902457894423614,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1777,
      "step": 33196
    },
    {
      "epoch": 7.902695947152294,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 33197
    },
    {
      "epoch": 7.902933999880974,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9875,
      "step": 33198
    },
    {
      "epoch": 7.903172052609653,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 33199
    },
    {
      "epoch": 7.903410105338333,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 33200
    },
    {
      "epoch": 7.903410105338333,
      "eval_runtime": 0.5304,
      "eval_samples_per_second": 1.885,
      "eval_steps_per_second": 1.885,
      "step": 33200
    },
    {
      "epoch": 7.9036481580670115,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 33201
    },
    {
      "epoch": 7.903886210795691,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 33202
    },
    {
      "epoch": 7.904124263524371,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 33203
    },
    {
      "epoch": 7.90436231625305,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9174,
      "step": 33204
    },
    {
      "epoch": 7.904600368981729,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 33205
    },
    {
      "epoch": 7.904838421710409,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 33206
    },
    {
      "epoch": 7.905076474439088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2951,
      "step": 33207
    },
    {
      "epoch": 7.905314527167768,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 33208
    },
    {
      "epoch": 7.905552579896447,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 33209
    },
    {
      "epoch": 7.905790632625126,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 33210
    },
    {
      "epoch": 7.906028685353806,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 33211
    },
    {
      "epoch": 7.906266738082485,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 33212
    },
    {
      "epoch": 7.906504790811165,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.3251,
      "step": 33213
    },
    {
      "epoch": 7.906742843539844,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 33214
    },
    {
      "epoch": 7.906980896268523,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 33215
    },
    {
      "epoch": 7.907218948997203,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 33216
    },
    {
      "epoch": 7.9074570017258825,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 33217
    },
    {
      "epoch": 7.907695054454562,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.6398,
      "step": 33218
    },
    {
      "epoch": 7.907933107183241,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.3253,
      "step": 33219
    },
    {
      "epoch": 7.90817115991192,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 33220
    },
    {
      "epoch": 7.9084092126406,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1553,
      "step": 33221
    },
    {
      "epoch": 7.90864726536928,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 33222
    },
    {
      "epoch": 7.908885318097958,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 33223
    },
    {
      "epoch": 7.909123370826638,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.7503,
      "step": 33224
    },
    {
      "epoch": 7.909361423555318,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 33225
    },
    {
      "epoch": 7.909599476283997,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 33226
    },
    {
      "epoch": 7.909837529012677,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 33227
    },
    {
      "epoch": 7.9100755817413555,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 33228
    },
    {
      "epoch": 7.910313634470035,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2618,
      "step": 33229
    },
    {
      "epoch": 7.910551687198715,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 33230
    },
    {
      "epoch": 7.910789739927393,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 33231
    },
    {
      "epoch": 7.911027792656073,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 33232
    },
    {
      "epoch": 7.911265845384753,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 33233
    },
    {
      "epoch": 7.911503898113432,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 33234
    },
    {
      "epoch": 7.911741950842112,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2621,
      "step": 33235
    },
    {
      "epoch": 7.911980003570791,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 33236
    },
    {
      "epoch": 7.91221805629947,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8145,
      "step": 33237
    },
    {
      "epoch": 7.91245610902815,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 33238
    },
    {
      "epoch": 7.912694161756829,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 33239
    },
    {
      "epoch": 7.912932214485508,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 33240
    },
    {
      "epoch": 7.913170267214188,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 33241
    },
    {
      "epoch": 7.913408319942867,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 33242
    },
    {
      "epoch": 7.913646372671547,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 33243
    },
    {
      "epoch": 7.9138844254002265,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 33244
    },
    {
      "epoch": 7.914122478128905,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 33245
    },
    {
      "epoch": 7.914360530857585,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 33246
    },
    {
      "epoch": 7.9145985835862644,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1851,
      "step": 33247
    },
    {
      "epoch": 7.914836636314944,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 33248
    },
    {
      "epoch": 7.915074689043623,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 33249
    },
    {
      "epoch": 7.915312741772302,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 33250
    },
    {
      "epoch": 7.915550794500982,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 33251
    },
    {
      "epoch": 7.915788847229662,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 33252
    },
    {
      "epoch": 7.916026899958341,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8668,
      "step": 33253
    },
    {
      "epoch": 7.91626495268702,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.912,
      "step": 33254
    },
    {
      "epoch": 7.9165030054156995,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 33255
    },
    {
      "epoch": 7.916741058144379,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7328,
      "step": 33256
    },
    {
      "epoch": 7.916979110873059,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 33257
    },
    {
      "epoch": 7.917217163601737,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 33258
    },
    {
      "epoch": 7.917455216330417,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 33259
    },
    {
      "epoch": 7.917693269059097,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.181,
      "step": 33260
    },
    {
      "epoch": 7.917931321787776,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 33261
    },
    {
      "epoch": 7.918169374516456,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 33262
    },
    {
      "epoch": 7.918407427245135,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 33263
    },
    {
      "epoch": 7.918645479973814,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9185,
      "step": 33264
    },
    {
      "epoch": 7.918883532702494,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 33265
    },
    {
      "epoch": 7.9191215854311725,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 33266
    },
    {
      "epoch": 7.919359638159852,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9862,
      "step": 33267
    },
    {
      "epoch": 7.919597690888532,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 33268
    },
    {
      "epoch": 7.919835743617211,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 33269
    },
    {
      "epoch": 7.920073796345891,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 33270
    },
    {
      "epoch": 7.92031184907457,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8973,
      "step": 33271
    },
    {
      "epoch": 7.920549901803249,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 33272
    },
    {
      "epoch": 7.920787954531929,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 33273
    },
    {
      "epoch": 7.9210260072606085,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.3902,
      "step": 33274
    },
    {
      "epoch": 7.921264059989287,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 33275
    },
    {
      "epoch": 7.921502112717967,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3036,
      "step": 33276
    },
    {
      "epoch": 7.921740165446646,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 33277
    },
    {
      "epoch": 7.921978218175326,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9786,
      "step": 33278
    },
    {
      "epoch": 7.922216270904006,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 33279
    },
    {
      "epoch": 7.922454323632684,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 33280
    },
    {
      "epoch": 7.922692376361364,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9776,
      "step": 33281
    },
    {
      "epoch": 7.9229304290900435,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 33282
    },
    {
      "epoch": 7.923168481818723,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 33283
    },
    {
      "epoch": 7.923406534547402,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 33284
    },
    {
      "epoch": 7.9236445872760815,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 33285
    },
    {
      "epoch": 7.923882640004761,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 33286
    },
    {
      "epoch": 7.924120692733441,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 33287
    },
    {
      "epoch": 7.92435874546212,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 33288
    },
    {
      "epoch": 7.924596798190799,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2593,
      "step": 33289
    },
    {
      "epoch": 7.924834850919479,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 33290
    },
    {
      "epoch": 7.925072903648158,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 33291
    },
    {
      "epoch": 7.925310956376838,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 33292
    },
    {
      "epoch": 7.9255490091055165,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 33293
    },
    {
      "epoch": 7.925787061834196,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 33294
    },
    {
      "epoch": 7.926025114562876,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9152,
      "step": 33295
    },
    {
      "epoch": 7.926263167291555,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 33296
    },
    {
      "epoch": 7.926501220020235,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 33297
    },
    {
      "epoch": 7.926739272748914,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0986,
      "step": 33298
    },
    {
      "epoch": 7.926977325477593,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 33299
    },
    {
      "epoch": 7.927215378206273,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 33300
    },
    {
      "epoch": 7.927215378206273,
      "eval_runtime": 0.5296,
      "eval_samples_per_second": 1.888,
      "eval_steps_per_second": 1.888,
      "step": 33300
    },
    {
      "epoch": 7.9274534309349525,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 33301
    },
    {
      "epoch": 7.927691483663631,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 33302
    },
    {
      "epoch": 7.927929536392311,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 33303
    },
    {
      "epoch": 7.92816758912099,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 33304
    },
    {
      "epoch": 7.92840564184967,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 33305
    },
    {
      "epoch": 7.928643694578349,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 33306
    },
    {
      "epoch": 7.928881747307028,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 33307
    },
    {
      "epoch": 7.929119800035708,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8667,
      "step": 33308
    },
    {
      "epoch": 7.9293578527643875,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 33309
    },
    {
      "epoch": 7.929595905493066,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 33310
    },
    {
      "epoch": 7.929833958221746,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1487,
      "step": 33311
    },
    {
      "epoch": 7.9300720109504255,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 33312
    },
    {
      "epoch": 7.930310063679105,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 33313
    },
    {
      "epoch": 7.930548116407785,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 33314
    },
    {
      "epoch": 7.930786169136463,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0087,
      "step": 33315
    },
    {
      "epoch": 7.931024221865143,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 33316
    },
    {
      "epoch": 7.931262274593823,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1691,
      "step": 33317
    },
    {
      "epoch": 7.931500327322502,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 33318
    },
    {
      "epoch": 7.931738380051181,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 33319
    },
    {
      "epoch": 7.9319764327798605,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 33320
    },
    {
      "epoch": 7.93221448550854,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 33321
    },
    {
      "epoch": 7.93245253823722,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 33322
    },
    {
      "epoch": 7.932690590965899,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 33323
    },
    {
      "epoch": 7.932928643694578,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 33324
    },
    {
      "epoch": 7.933166696423258,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4274,
      "step": 33325
    },
    {
      "epoch": 7.933404749151937,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 33326
    },
    {
      "epoch": 7.933642801880617,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 33327
    },
    {
      "epoch": 7.933880854609296,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 33328
    },
    {
      "epoch": 7.934118907337975,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 33329
    },
    {
      "epoch": 7.934356960066655,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 33330
    },
    {
      "epoch": 7.934595012795334,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2358,
      "step": 33331
    },
    {
      "epoch": 7.934833065524014,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.972,
      "step": 33332
    },
    {
      "epoch": 7.935071118252693,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1988,
      "step": 33333
    },
    {
      "epoch": 7.935309170981372,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.5195,
      "step": 33334
    },
    {
      "epoch": 7.935547223710052,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.6419,
      "step": 33335
    },
    {
      "epoch": 7.9357852764387316,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 33336
    },
    {
      "epoch": 7.93602332916741,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 33337
    },
    {
      "epoch": 7.93626138189609,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 33338
    },
    {
      "epoch": 7.9364994346247695,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 33339
    },
    {
      "epoch": 7.936737487353449,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 33340
    },
    {
      "epoch": 7.936975540082129,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 33341
    },
    {
      "epoch": 7.937213592810807,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4621,
      "step": 33342
    },
    {
      "epoch": 7.937451645539487,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 33343
    },
    {
      "epoch": 7.937689698268167,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 33344
    },
    {
      "epoch": 7.937927750996845,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 33345
    },
    {
      "epoch": 7.938165803725525,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1212,
      "step": 33346
    },
    {
      "epoch": 7.9384038564542045,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 33347
    },
    {
      "epoch": 7.938641909182884,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 33348
    },
    {
      "epoch": 7.938879961911564,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 33349
    },
    {
      "epoch": 7.9391180146402425,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 33350
    },
    {
      "epoch": 7.939356067368922,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 33351
    },
    {
      "epoch": 7.939594120097602,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1714,
      "step": 33352
    },
    {
      "epoch": 7.939832172826281,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 33353
    },
    {
      "epoch": 7.94007022555496,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 33354
    },
    {
      "epoch": 7.94030827828364,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 33355
    },
    {
      "epoch": 7.940546331012319,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 33356
    },
    {
      "epoch": 7.940784383740999,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 33357
    },
    {
      "epoch": 7.941022436469678,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 33358
    },
    {
      "epoch": 7.941260489198357,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1124,
      "step": 33359
    },
    {
      "epoch": 7.941498541927037,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 33360
    },
    {
      "epoch": 7.941736594655716,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 33361
    },
    {
      "epoch": 7.941974647384396,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 33362
    },
    {
      "epoch": 7.942212700113075,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 33363
    },
    {
      "epoch": 7.942450752841754,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 33364
    },
    {
      "epoch": 7.942688805570434,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3836,
      "step": 33365
    },
    {
      "epoch": 7.9429268582991135,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0615,
      "step": 33366
    },
    {
      "epoch": 7.943164911027793,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 33367
    },
    {
      "epoch": 7.943402963756472,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 33368
    },
    {
      "epoch": 7.943641016485151,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 33369
    },
    {
      "epoch": 7.943879069213831,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 33370
    },
    {
      "epoch": 7.944117121942511,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2101,
      "step": 33371
    },
    {
      "epoch": 7.944355174671189,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.338,
      "step": 33372
    },
    {
      "epoch": 7.944593227399869,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 33373
    },
    {
      "epoch": 7.944831280128549,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 33374
    },
    {
      "epoch": 7.945069332857228,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9893,
      "step": 33375
    },
    {
      "epoch": 7.945307385585908,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 33376
    },
    {
      "epoch": 7.9455454383145865,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 33377
    },
    {
      "epoch": 7.945783491043266,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 33378
    },
    {
      "epoch": 7.946021543771946,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 33379
    },
    {
      "epoch": 7.946259596500624,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1243,
      "step": 33380
    },
    {
      "epoch": 7.946497649229304,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1431,
      "step": 33381
    },
    {
      "epoch": 7.946735701957984,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 33382
    },
    {
      "epoch": 7.946973754686663,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 33383
    },
    {
      "epoch": 7.947211807415343,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4318,
      "step": 33384
    },
    {
      "epoch": 7.9474498601440215,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 33385
    },
    {
      "epoch": 7.947687912872701,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1175,
      "step": 33386
    },
    {
      "epoch": 7.947925965601381,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 33387
    },
    {
      "epoch": 7.94816401833006,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 33388
    },
    {
      "epoch": 7.948402071058739,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.893,
      "step": 33389
    },
    {
      "epoch": 7.948640123787419,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 33390
    },
    {
      "epoch": 7.948878176516098,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.4173,
      "step": 33391
    },
    {
      "epoch": 7.949116229244778,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 33392
    },
    {
      "epoch": 7.9493542819734575,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 33393
    },
    {
      "epoch": 7.949592334702136,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 33394
    },
    {
      "epoch": 7.949830387430816,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8108,
      "step": 33395
    },
    {
      "epoch": 7.950068440159495,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 33396
    },
    {
      "epoch": 7.950306492888175,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 33397
    },
    {
      "epoch": 7.950544545616854,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 33398
    },
    {
      "epoch": 7.950782598345533,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 33399
    },
    {
      "epoch": 7.951020651074213,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 33400
    },
    {
      "epoch": 7.951020651074213,
      "eval_runtime": 0.5155,
      "eval_samples_per_second": 1.94,
      "eval_steps_per_second": 1.94,
      "step": 33400
    },
    {
      "epoch": 7.951258703802893,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 33401
    },
    {
      "epoch": 7.951496756531572,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 33402
    },
    {
      "epoch": 7.951734809260251,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 33403
    },
    {
      "epoch": 7.9519728619889305,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.8177,
      "step": 33404
    },
    {
      "epoch": 7.95221091471761,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8836,
      "step": 33405
    },
    {
      "epoch": 7.95244896744629,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 33406
    },
    {
      "epoch": 7.952687020174968,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3534,
      "step": 33407
    },
    {
      "epoch": 7.952925072903648,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 33408
    },
    {
      "epoch": 7.953163125632328,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 33409
    },
    {
      "epoch": 7.953401178361007,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 33410
    },
    {
      "epoch": 7.953639231089687,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 33411
    },
    {
      "epoch": 7.953877283818366,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 33412
    },
    {
      "epoch": 7.954115336547045,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 33413
    },
    {
      "epoch": 7.954353389275725,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.6657,
      "step": 33414
    },
    {
      "epoch": 7.9545914420044035,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1061,
      "step": 33415
    },
    {
      "epoch": 7.954829494733083,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1768,
      "step": 33416
    },
    {
      "epoch": 7.955067547461763,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8292,
      "step": 33417
    },
    {
      "epoch": 7.955305600190442,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 33418
    },
    {
      "epoch": 7.955543652919122,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 33419
    },
    {
      "epoch": 7.955781705647801,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 33420
    },
    {
      "epoch": 7.95601975837648,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8842,
      "step": 33421
    },
    {
      "epoch": 7.95625781110516,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 33422
    },
    {
      "epoch": 7.956495863833839,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 33423
    },
    {
      "epoch": 7.956733916562518,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 33424
    },
    {
      "epoch": 7.956971969291198,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 33425
    },
    {
      "epoch": 7.957210022019877,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2316,
      "step": 33426
    },
    {
      "epoch": 7.957448074748557,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 33427
    },
    {
      "epoch": 7.957686127477237,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 33428
    },
    {
      "epoch": 7.957924180205915,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0628,
      "step": 33429
    },
    {
      "epoch": 7.958162232934595,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0455,
      "step": 33430
    },
    {
      "epoch": 7.9584002856632745,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1595,
      "step": 33431
    },
    {
      "epoch": 7.958638338391954,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 33432
    },
    {
      "epoch": 7.958876391120633,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 33433
    },
    {
      "epoch": 7.959114443849312,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1319,
      "step": 33434
    },
    {
      "epoch": 7.959352496577992,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2017,
      "step": 33435
    },
    {
      "epoch": 7.959590549306672,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 33436
    },
    {
      "epoch": 7.959828602035351,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 33437
    },
    {
      "epoch": 7.96006665476403,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 33438
    },
    {
      "epoch": 7.96030470749271,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.14,
      "step": 33439
    },
    {
      "epoch": 7.960542760221389,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 33440
    },
    {
      "epoch": 7.960780812950069,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3601,
      "step": 33441
    },
    {
      "epoch": 7.9610188656787475,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 33442
    },
    {
      "epoch": 7.961256918407427,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 33443
    },
    {
      "epoch": 7.961494971136107,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 33444
    },
    {
      "epoch": 7.961733023864786,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4095,
      "step": 33445
    },
    {
      "epoch": 7.961971076593466,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 33446
    },
    {
      "epoch": 7.962209129322145,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 33447
    },
    {
      "epoch": 7.962447182050824,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 33448
    },
    {
      "epoch": 7.962685234779504,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 33449
    },
    {
      "epoch": 7.962923287508183,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 33450
    },
    {
      "epoch": 7.963161340236862,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8105,
      "step": 33451
    },
    {
      "epoch": 7.963399392965542,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 33452
    },
    {
      "epoch": 7.963637445694221,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0331,
      "step": 33453
    },
    {
      "epoch": 7.963875498422901,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 33454
    },
    {
      "epoch": 7.96411355115158,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1362,
      "step": 33455
    },
    {
      "epoch": 7.964351603880259,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 33456
    },
    {
      "epoch": 7.964589656608939,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 33457
    },
    {
      "epoch": 7.9648277093376185,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 33458
    },
    {
      "epoch": 7.965065762066297,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 33459
    },
    {
      "epoch": 7.965303814794977,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 33460
    },
    {
      "epoch": 7.9655418675236564,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1763,
      "step": 33461
    },
    {
      "epoch": 7.965779920252336,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 33462
    },
    {
      "epoch": 7.966017972981016,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 33463
    },
    {
      "epoch": 7.966256025709694,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1724,
      "step": 33464
    },
    {
      "epoch": 7.966494078438374,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.873,
      "step": 33465
    },
    {
      "epoch": 7.966732131167054,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 33466
    },
    {
      "epoch": 7.966970183895733,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9476,
      "step": 33467
    },
    {
      "epoch": 7.967208236624412,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8355,
      "step": 33468
    },
    {
      "epoch": 7.9674462893530915,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 33469
    },
    {
      "epoch": 7.967684342081771,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 33470
    },
    {
      "epoch": 7.967922394810451,
      "grad_norm": 8.1875,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 33471
    },
    {
      "epoch": 7.96816044753913,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 33472
    },
    {
      "epoch": 7.968398500267809,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.2095,
      "step": 33473
    },
    {
      "epoch": 7.968636552996489,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 33474
    },
    {
      "epoch": 7.968874605725168,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1479,
      "step": 33475
    },
    {
      "epoch": 7.969112658453848,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4337,
      "step": 33476
    },
    {
      "epoch": 7.969350711182527,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 33477
    },
    {
      "epoch": 7.969588763911206,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1461,
      "step": 33478
    },
    {
      "epoch": 7.969826816639886,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1132,
      "step": 33479
    },
    {
      "epoch": 7.970064869368565,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 33480
    },
    {
      "epoch": 7.970302922097245,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0938,
      "step": 33481
    },
    {
      "epoch": 7.970540974825924,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 33482
    },
    {
      "epoch": 7.970779027554603,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 33483
    },
    {
      "epoch": 7.971017080283283,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 33484
    },
    {
      "epoch": 7.9712551330119625,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 33485
    },
    {
      "epoch": 7.971493185740641,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 33486
    },
    {
      "epoch": 7.971731238469321,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3144,
      "step": 33487
    },
    {
      "epoch": 7.9719692911980005,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 33488
    },
    {
      "epoch": 7.97220734392668,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 33489
    },
    {
      "epoch": 7.972445396655359,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 33490
    },
    {
      "epoch": 7.972683449384038,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 33491
    },
    {
      "epoch": 7.972921502112718,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2075,
      "step": 33492
    },
    {
      "epoch": 7.973159554841398,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 33493
    },
    {
      "epoch": 7.973397607570076,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 33494
    },
    {
      "epoch": 7.973635660298756,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0802,
      "step": 33495
    },
    {
      "epoch": 7.9738737130274355,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 33496
    },
    {
      "epoch": 7.974111765756115,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2465,
      "step": 33497
    },
    {
      "epoch": 7.974349818484795,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2391,
      "step": 33498
    },
    {
      "epoch": 7.9745878712134735,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3284,
      "step": 33499
    },
    {
      "epoch": 7.974825923942153,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 33500
    },
    {
      "epoch": 7.974825923942153,
      "eval_runtime": 0.6059,
      "eval_samples_per_second": 1.65,
      "eval_steps_per_second": 1.65,
      "step": 33500
    },
    {
      "epoch": 7.975063976670833,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9674,
      "step": 33501
    },
    {
      "epoch": 7.975302029399512,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1594,
      "step": 33502
    },
    {
      "epoch": 7.975540082128191,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 33503
    },
    {
      "epoch": 7.975778134856871,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 33504
    },
    {
      "epoch": 7.97601618758555,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 33505
    },
    {
      "epoch": 7.97625424031423,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0868,
      "step": 33506
    },
    {
      "epoch": 7.976492293042909,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.086,
      "step": 33507
    },
    {
      "epoch": 7.976730345771588,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 33508
    },
    {
      "epoch": 7.976968398500268,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 33509
    },
    {
      "epoch": 7.977206451228947,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1901,
      "step": 33510
    },
    {
      "epoch": 7.977444503957627,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 33511
    },
    {
      "epoch": 7.977682556686306,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 33512
    },
    {
      "epoch": 7.977920609414985,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9006,
      "step": 33513
    },
    {
      "epoch": 7.978158662143665,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 33514
    },
    {
      "epoch": 7.9783967148723445,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 33515
    },
    {
      "epoch": 7.978634767601024,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 33516
    },
    {
      "epoch": 7.978872820329703,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 33517
    },
    {
      "epoch": 7.979110873058382,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 33518
    },
    {
      "epoch": 7.979348925787062,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 33519
    },
    {
      "epoch": 7.979586978515742,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.4291,
      "step": 33520
    },
    {
      "epoch": 7.97982503124442,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 33521
    },
    {
      "epoch": 7.9800630839731,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 33522
    },
    {
      "epoch": 7.9803011367017795,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 33523
    },
    {
      "epoch": 7.980539189430459,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 33524
    },
    {
      "epoch": 7.980777242159138,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9384,
      "step": 33525
    },
    {
      "epoch": 7.9810152948878175,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 33526
    },
    {
      "epoch": 7.981253347616497,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.5246,
      "step": 33527
    },
    {
      "epoch": 7.981491400345177,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2964,
      "step": 33528
    },
    {
      "epoch": 7.981729453073855,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3406,
      "step": 33529
    },
    {
      "epoch": 7.981967505802535,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2145,
      "step": 33530
    },
    {
      "epoch": 7.982205558531215,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0812,
      "step": 33531
    },
    {
      "epoch": 7.982443611259894,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.4358,
      "step": 33532
    },
    {
      "epoch": 7.982681663988574,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 33533
    },
    {
      "epoch": 7.9829197167172525,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 33534
    },
    {
      "epoch": 7.983157769445932,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3514,
      "step": 33535
    },
    {
      "epoch": 7.983395822174612,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.787,
      "step": 33536
    },
    {
      "epoch": 7.983633874903291,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3242,
      "step": 33537
    },
    {
      "epoch": 7.98387192763197,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0854,
      "step": 33538
    },
    {
      "epoch": 7.98410998036065,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 33539
    },
    {
      "epoch": 7.984348033089329,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2231,
      "step": 33540
    },
    {
      "epoch": 7.984586085818009,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 33541
    },
    {
      "epoch": 7.9848241385466885,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.657,
      "step": 33542
    },
    {
      "epoch": 7.985062191275367,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 33543
    },
    {
      "epoch": 7.985300244004047,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7001,
      "step": 33544
    },
    {
      "epoch": 7.985538296732726,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 33545
    },
    {
      "epoch": 7.985776349461406,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 33546
    },
    {
      "epoch": 7.986014402190085,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1731,
      "step": 33547
    },
    {
      "epoch": 7.986252454918764,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0485,
      "step": 33548
    },
    {
      "epoch": 7.986490507647444,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 33549
    },
    {
      "epoch": 7.9867285603761236,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 33550
    },
    {
      "epoch": 7.986966613104803,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 33551
    },
    {
      "epoch": 7.987204665833482,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 33552
    },
    {
      "epoch": 7.9874427185621615,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2471,
      "step": 33553
    },
    {
      "epoch": 7.987680771290841,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 33554
    },
    {
      "epoch": 7.987918824019521,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 33555
    },
    {
      "epoch": 7.988156876748199,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 33556
    },
    {
      "epoch": 7.988394929476879,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 33557
    },
    {
      "epoch": 7.988632982205559,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 33558
    },
    {
      "epoch": 7.988871034934238,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 33559
    },
    {
      "epoch": 7.989109087662918,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 33560
    },
    {
      "epoch": 7.9893471403915965,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 33561
    },
    {
      "epoch": 7.989585193120276,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 33562
    },
    {
      "epoch": 7.989823245848956,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 33563
    },
    {
      "epoch": 7.9900612985776345,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 33564
    },
    {
      "epoch": 7.990299351306314,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 33565
    },
    {
      "epoch": 7.990537404034994,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 33566
    },
    {
      "epoch": 7.990775456763673,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 33567
    },
    {
      "epoch": 7.991013509492353,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 33568
    },
    {
      "epoch": 7.991251562221032,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9853,
      "step": 33569
    },
    {
      "epoch": 7.991489614949711,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4622,
      "step": 33570
    },
    {
      "epoch": 7.991727667678391,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2338,
      "step": 33571
    },
    {
      "epoch": 7.99196572040707,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7501,
      "step": 33572
    },
    {
      "epoch": 7.992203773135749,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 33573
    },
    {
      "epoch": 7.992441825864429,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 33574
    },
    {
      "epoch": 7.992679878593108,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 33575
    },
    {
      "epoch": 7.992917931321788,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1797,
      "step": 33576
    },
    {
      "epoch": 7.993155984050468,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2446,
      "step": 33577
    },
    {
      "epoch": 7.993394036779146,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9354,
      "step": 33578
    },
    {
      "epoch": 7.993632089507826,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1899,
      "step": 33579
    },
    {
      "epoch": 7.9938701422365055,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 33580
    },
    {
      "epoch": 7.994108194965185,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4155,
      "step": 33581
    },
    {
      "epoch": 7.994346247693864,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 33582
    },
    {
      "epoch": 7.994584300422543,
      "grad_norm": 2.75,
      "learning_rate": 3e-05,
      "loss": 1.7738,
      "step": 33583
    },
    {
      "epoch": 7.994822353151223,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 33584
    },
    {
      "epoch": 7.995060405879903,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 33585
    },
    {
      "epoch": 7.995298458608582,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 33586
    },
    {
      "epoch": 7.995536511337261,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 33587
    },
    {
      "epoch": 7.995774564065941,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 33588
    },
    {
      "epoch": 7.99601261679462,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 33589
    },
    {
      "epoch": 7.9962506695233,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7878,
      "step": 33590
    },
    {
      "epoch": 7.9964887222519785,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 33591
    },
    {
      "epoch": 7.996726774980658,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.028,
      "step": 33592
    },
    {
      "epoch": 7.996964827709338,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 33593
    },
    {
      "epoch": 7.997202880438017,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 33594
    },
    {
      "epoch": 7.997440933166697,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 33595
    },
    {
      "epoch": 7.997678985895376,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 33596
    },
    {
      "epoch": 7.997917038624055,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9512,
      "step": 33597
    },
    {
      "epoch": 7.998155091352735,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1934,
      "step": 33598
    },
    {
      "epoch": 7.9983931440814136,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1904,
      "step": 33599
    },
    {
      "epoch": 7.998631196810093,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 33600
    },
    {
      "epoch": 7.998631196810093,
      "eval_runtime": 0.5248,
      "eval_samples_per_second": 1.906,
      "eval_steps_per_second": 1.906,
      "step": 33600
    },
    {
      "epoch": 7.998869249538773,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 33601
    },
    {
      "epoch": 7.999107302267452,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 33602
    },
    {
      "epoch": 7.999345354996132,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 33603
    },
    {
      "epoch": 7.999583407724811,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 33604
    },
    {
      "epoch": 7.99982146045349,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 33605
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 33606
    },
    {
      "epoch": 8.00023805272868,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 33607
    },
    {
      "epoch": 8.00047610545736,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 33608
    },
    {
      "epoch": 8.000714158186039,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 33609
    },
    {
      "epoch": 8.000952210914718,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 33610
    },
    {
      "epoch": 8.001190263643396,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 33611
    },
    {
      "epoch": 8.001428316372076,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 33612
    },
    {
      "epoch": 8.001666369100755,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 33613
    },
    {
      "epoch": 8.001904421829435,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8435,
      "step": 33614
    },
    {
      "epoch": 8.002142474558115,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0153,
      "step": 33615
    },
    {
      "epoch": 8.002380527286794,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 33616
    },
    {
      "epoch": 8.002618580015474,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 33617
    },
    {
      "epoch": 8.002856632744153,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 33618
    },
    {
      "epoch": 8.003094685472833,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 33619
    },
    {
      "epoch": 8.003332738201511,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9062,
      "step": 33620
    },
    {
      "epoch": 8.00357079093019,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 33621
    },
    {
      "epoch": 8.00380884365887,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 33622
    },
    {
      "epoch": 8.00404689638755,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 33623
    },
    {
      "epoch": 8.00428494911623,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 33624
    },
    {
      "epoch": 8.004523001844909,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8451,
      "step": 33625
    },
    {
      "epoch": 8.004761054573589,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 33626
    },
    {
      "epoch": 8.004999107302268,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 33627
    },
    {
      "epoch": 8.005237160030946,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 33628
    },
    {
      "epoch": 8.005475212759626,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.151,
      "step": 33629
    },
    {
      "epoch": 8.005713265488305,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 33630
    },
    {
      "epoch": 8.005951318216985,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 33631
    },
    {
      "epoch": 8.006189370945664,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 33632
    },
    {
      "epoch": 8.006427423674344,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 33633
    },
    {
      "epoch": 8.006665476403024,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 33634
    },
    {
      "epoch": 8.006903529131703,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 33635
    },
    {
      "epoch": 8.007141581860383,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 33636
    },
    {
      "epoch": 8.00737963458906,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 33637
    },
    {
      "epoch": 8.00761768731774,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 33638
    },
    {
      "epoch": 8.00785574004642,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 33639
    },
    {
      "epoch": 8.0080937927751,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 33640
    },
    {
      "epoch": 8.008331845503779,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.755,
      "step": 33641
    },
    {
      "epoch": 8.008569898232459,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 33642
    },
    {
      "epoch": 8.008807950961138,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 33643
    },
    {
      "epoch": 8.009046003689818,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 33644
    },
    {
      "epoch": 8.009284056418497,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 33645
    },
    {
      "epoch": 8.009522109147175,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3708,
      "step": 33646
    },
    {
      "epoch": 8.009760161875855,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 33647
    },
    {
      "epoch": 8.009998214604535,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 33648
    },
    {
      "epoch": 8.010236267333214,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1274,
      "step": 33649
    },
    {
      "epoch": 8.010474320061894,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 33650
    },
    {
      "epoch": 8.010712372790573,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8883,
      "step": 33651
    },
    {
      "epoch": 8.010950425519253,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 33652
    },
    {
      "epoch": 8.011188478247933,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 33653
    },
    {
      "epoch": 8.011426530976612,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 33654
    },
    {
      "epoch": 8.01166458370529,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 33655
    },
    {
      "epoch": 8.01190263643397,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 33656
    },
    {
      "epoch": 8.01214068916265,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 33657
    },
    {
      "epoch": 8.012378741891329,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 33658
    },
    {
      "epoch": 8.012616794620008,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0745,
      "step": 33659
    },
    {
      "epoch": 8.012854847348688,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 33660
    },
    {
      "epoch": 8.013092900077368,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9801,
      "step": 33661
    },
    {
      "epoch": 8.013330952806047,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 33662
    },
    {
      "epoch": 8.013569005534725,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 33663
    },
    {
      "epoch": 8.013807058263405,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 33664
    },
    {
      "epoch": 8.014045110992084,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 33665
    },
    {
      "epoch": 8.014283163720764,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 33666
    },
    {
      "epoch": 8.014521216449443,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9486,
      "step": 33667
    },
    {
      "epoch": 8.014759269178123,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7602,
      "step": 33668
    },
    {
      "epoch": 8.014997321906803,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 33669
    },
    {
      "epoch": 8.015235374635482,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.2051,
      "step": 33670
    },
    {
      "epoch": 8.015473427364162,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.741,
      "step": 33671
    },
    {
      "epoch": 8.01571148009284,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0573,
      "step": 33672
    },
    {
      "epoch": 8.01594953282152,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 33673
    },
    {
      "epoch": 8.016187585550199,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 33674
    },
    {
      "epoch": 8.016425638278879,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3405,
      "step": 33675
    },
    {
      "epoch": 8.016663691007558,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 33676
    },
    {
      "epoch": 8.016901743736238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3516,
      "step": 33677
    },
    {
      "epoch": 8.017139796464917,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 33678
    },
    {
      "epoch": 8.017377849193597,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3572,
      "step": 33679
    },
    {
      "epoch": 8.017615901922277,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9807,
      "step": 33680
    },
    {
      "epoch": 8.017853954650954,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 33681
    },
    {
      "epoch": 8.018092007379634,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 33682
    },
    {
      "epoch": 8.018330060108314,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8738,
      "step": 33683
    },
    {
      "epoch": 8.018568112836993,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 33684
    },
    {
      "epoch": 8.018806165565673,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 33685
    },
    {
      "epoch": 8.019044218294352,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 33686
    },
    {
      "epoch": 8.019282271023032,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 33687
    },
    {
      "epoch": 8.019520323751712,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 33688
    },
    {
      "epoch": 8.019758376480391,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 33689
    },
    {
      "epoch": 8.019996429209069,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 33690
    },
    {
      "epoch": 8.020234481937749,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 33691
    },
    {
      "epoch": 8.020472534666428,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 33692
    },
    {
      "epoch": 8.020710587395108,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1131,
      "step": 33693
    },
    {
      "epoch": 8.020948640123787,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8688,
      "step": 33694
    },
    {
      "epoch": 8.021186692852467,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 33695
    },
    {
      "epoch": 8.021424745581147,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 33696
    },
    {
      "epoch": 8.021662798309826,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 33697
    },
    {
      "epoch": 8.021900851038504,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 33698
    },
    {
      "epoch": 8.022138903767184,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 33699
    },
    {
      "epoch": 8.022376956495863,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 33700
    },
    {
      "epoch": 8.022376956495863,
      "eval_runtime": 0.5279,
      "eval_samples_per_second": 1.894,
      "eval_steps_per_second": 1.894,
      "step": 33700
    },
    {
      "epoch": 8.022615009224543,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 33701
    },
    {
      "epoch": 8.022853061953223,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.6653,
      "step": 33702
    },
    {
      "epoch": 8.023091114681902,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 33703
    },
    {
      "epoch": 8.023329167410582,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 33704
    },
    {
      "epoch": 8.023567220139261,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 33705
    },
    {
      "epoch": 8.023805272867941,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.861,
      "step": 33706
    },
    {
      "epoch": 8.024043325596619,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7452,
      "step": 33707
    },
    {
      "epoch": 8.024281378325298,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8436,
      "step": 33708
    },
    {
      "epoch": 8.024519431053978,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 33709
    },
    {
      "epoch": 8.024757483782658,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 33710
    },
    {
      "epoch": 8.024995536511337,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 33711
    },
    {
      "epoch": 8.025233589240017,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 33712
    },
    {
      "epoch": 8.025471641968696,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.1623,
      "step": 33713
    },
    {
      "epoch": 8.025709694697376,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 33714
    },
    {
      "epoch": 8.025947747426056,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 33715
    },
    {
      "epoch": 8.026185800154733,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8916,
      "step": 33716
    },
    {
      "epoch": 8.026423852883413,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 33717
    },
    {
      "epoch": 8.026661905612093,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1738,
      "step": 33718
    },
    {
      "epoch": 8.026899958340772,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 33719
    },
    {
      "epoch": 8.027138011069452,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 33720
    },
    {
      "epoch": 8.027376063798132,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 33721
    },
    {
      "epoch": 8.027614116526811,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1449,
      "step": 33722
    },
    {
      "epoch": 8.02785216925549,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 33723
    },
    {
      "epoch": 8.02809022198417,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 33724
    },
    {
      "epoch": 8.028328274712848,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 33725
    },
    {
      "epoch": 8.028566327441528,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 33726
    },
    {
      "epoch": 8.028804380170207,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 33727
    },
    {
      "epoch": 8.029042432898887,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 33728
    },
    {
      "epoch": 8.029280485627567,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1153,
      "step": 33729
    },
    {
      "epoch": 8.029518538356246,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 33730
    },
    {
      "epoch": 8.029756591084926,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2212,
      "step": 33731
    },
    {
      "epoch": 8.029994643813605,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 33732
    },
    {
      "epoch": 8.030232696542285,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 33733
    },
    {
      "epoch": 8.030470749270963,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 33734
    },
    {
      "epoch": 8.030708801999642,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7384,
      "step": 33735
    },
    {
      "epoch": 8.030946854728322,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.097,
      "step": 33736
    },
    {
      "epoch": 8.031184907457002,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 33737
    },
    {
      "epoch": 8.031422960185681,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 33738
    },
    {
      "epoch": 8.03166101291436,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 33739
    },
    {
      "epoch": 8.03189906564304,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 33740
    },
    {
      "epoch": 8.03213711837172,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 33741
    },
    {
      "epoch": 8.032375171100398,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 33742
    },
    {
      "epoch": 8.032613223829077,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 33743
    },
    {
      "epoch": 8.032851276557757,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2388,
      "step": 33744
    },
    {
      "epoch": 8.033089329286437,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 33745
    },
    {
      "epoch": 8.033327382015116,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2164,
      "step": 33746
    },
    {
      "epoch": 8.033565434743796,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 33747
    },
    {
      "epoch": 8.033803487472476,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 33748
    },
    {
      "epoch": 8.034041540201155,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 33749
    },
    {
      "epoch": 8.034279592929835,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 33750
    },
    {
      "epoch": 8.034517645658513,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 33751
    },
    {
      "epoch": 8.034755698387192,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 33752
    },
    {
      "epoch": 8.034993751115872,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 33753
    },
    {
      "epoch": 8.035231803844551,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3235,
      "step": 33754
    },
    {
      "epoch": 8.035469856573231,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3358,
      "step": 33755
    },
    {
      "epoch": 8.03570790930191,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 33756
    },
    {
      "epoch": 8.03594596203059,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 33757
    },
    {
      "epoch": 8.03618401475927,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 33758
    },
    {
      "epoch": 8.03642206748795,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.255,
      "step": 33759
    },
    {
      "epoch": 8.036660120216627,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3099,
      "step": 33760
    },
    {
      "epoch": 8.036898172945307,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 33761
    },
    {
      "epoch": 8.037136225673986,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.4307,
      "step": 33762
    },
    {
      "epoch": 8.037374278402666,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 33763
    },
    {
      "epoch": 8.037612331131346,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 33764
    },
    {
      "epoch": 8.037850383860025,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 33765
    },
    {
      "epoch": 8.038088436588705,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 33766
    },
    {
      "epoch": 8.038326489317384,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 33767
    },
    {
      "epoch": 8.038564542046064,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 33768
    },
    {
      "epoch": 8.038802594774742,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 33769
    },
    {
      "epoch": 8.039040647503422,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8712,
      "step": 33770
    },
    {
      "epoch": 8.039278700232101,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 33771
    },
    {
      "epoch": 8.03951675296078,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 33772
    },
    {
      "epoch": 8.03975480568946,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 33773
    },
    {
      "epoch": 8.03999285841814,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2777,
      "step": 33774
    },
    {
      "epoch": 8.04023091114682,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9009,
      "step": 33775
    },
    {
      "epoch": 8.0404689638755,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 33776
    },
    {
      "epoch": 8.040707016604177,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 33777
    },
    {
      "epoch": 8.040945069332857,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 33778
    },
    {
      "epoch": 8.041183122061536,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 33779
    },
    {
      "epoch": 8.041421174790216,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2404,
      "step": 33780
    },
    {
      "epoch": 8.041659227518895,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 33781
    },
    {
      "epoch": 8.041897280247575,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 33782
    },
    {
      "epoch": 8.042135332976255,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 33783
    },
    {
      "epoch": 8.042373385704934,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2802,
      "step": 33784
    },
    {
      "epoch": 8.042611438433614,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.7712,
      "step": 33785
    },
    {
      "epoch": 8.042849491162292,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.864,
      "step": 33786
    },
    {
      "epoch": 8.043087543890971,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0018,
      "step": 33787
    },
    {
      "epoch": 8.04332559661965,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 33788
    },
    {
      "epoch": 8.04356364934833,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 33789
    },
    {
      "epoch": 8.04380170207701,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.2837,
      "step": 33790
    },
    {
      "epoch": 8.04403975480569,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 33791
    },
    {
      "epoch": 8.04427780753437,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7724,
      "step": 33792
    },
    {
      "epoch": 8.044515860263049,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.816,
      "step": 33793
    },
    {
      "epoch": 8.044753912991728,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8493,
      "step": 33794
    },
    {
      "epoch": 8.044991965720406,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 33795
    },
    {
      "epoch": 8.045230018449086,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 33796
    },
    {
      "epoch": 8.045468071177766,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 33797
    },
    {
      "epoch": 8.045706123906445,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 33798
    },
    {
      "epoch": 8.045944176635125,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7949,
      "step": 33799
    },
    {
      "epoch": 8.046182229363804,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 33800
    },
    {
      "epoch": 8.046182229363804,
      "eval_runtime": 0.5404,
      "eval_samples_per_second": 1.85,
      "eval_steps_per_second": 1.85,
      "step": 33800
    },
    {
      "epoch": 8.046420282092484,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 33801
    },
    {
      "epoch": 8.046658334821164,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 33802
    },
    {
      "epoch": 8.046896387549843,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 33803
    },
    {
      "epoch": 8.047134440278521,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 33804
    },
    {
      "epoch": 8.0473724930072,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 33805
    },
    {
      "epoch": 8.04761054573588,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 33806
    },
    {
      "epoch": 8.04784859846456,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 33807
    },
    {
      "epoch": 8.04808665119324,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 33808
    },
    {
      "epoch": 8.048324703921919,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8699,
      "step": 33809
    },
    {
      "epoch": 8.048562756650599,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7789,
      "step": 33810
    },
    {
      "epoch": 8.048800809379278,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 33811
    },
    {
      "epoch": 8.049038862107956,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 33812
    },
    {
      "epoch": 8.049276914836636,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 33813
    },
    {
      "epoch": 8.049514967565315,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 33814
    },
    {
      "epoch": 8.049753020293995,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 33815
    },
    {
      "epoch": 8.049991073022674,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 33816
    },
    {
      "epoch": 8.050229125751354,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 33817
    },
    {
      "epoch": 8.050467178480034,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 33818
    },
    {
      "epoch": 8.050705231208713,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 33819
    },
    {
      "epoch": 8.050943283937393,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 33820
    },
    {
      "epoch": 8.05118133666607,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1639,
      "step": 33821
    },
    {
      "epoch": 8.05141938939475,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 33822
    },
    {
      "epoch": 8.05165744212343,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 33823
    },
    {
      "epoch": 8.05189549485211,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 33824
    },
    {
      "epoch": 8.05213354758079,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8786,
      "step": 33825
    },
    {
      "epoch": 8.052371600309469,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 33826
    },
    {
      "epoch": 8.052609653038148,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 33827
    },
    {
      "epoch": 8.052847705766828,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9073,
      "step": 33828
    },
    {
      "epoch": 8.053085758495508,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 33829
    },
    {
      "epoch": 8.053323811224185,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8251,
      "step": 33830
    },
    {
      "epoch": 8.053561863952865,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 33831
    },
    {
      "epoch": 8.053799916681545,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0952,
      "step": 33832
    },
    {
      "epoch": 8.054037969410224,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 33833
    },
    {
      "epoch": 8.054276022138904,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 33834
    },
    {
      "epoch": 8.054514074867583,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 33835
    },
    {
      "epoch": 8.054752127596263,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 33836
    },
    {
      "epoch": 8.054990180324943,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8938,
      "step": 33837
    },
    {
      "epoch": 8.055228233053622,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 33838
    },
    {
      "epoch": 8.0554662857823,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.818,
      "step": 33839
    },
    {
      "epoch": 8.05570433851098,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 33840
    },
    {
      "epoch": 8.05594239123966,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 33841
    },
    {
      "epoch": 8.056180443968339,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 33842
    },
    {
      "epoch": 8.056418496697018,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 33843
    },
    {
      "epoch": 8.056656549425698,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 33844
    },
    {
      "epoch": 8.056894602154378,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 33845
    },
    {
      "epoch": 8.057132654883057,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3129,
      "step": 33846
    },
    {
      "epoch": 8.057370707611735,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 33847
    },
    {
      "epoch": 8.057608760340415,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 33848
    },
    {
      "epoch": 8.057846813069094,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 33849
    },
    {
      "epoch": 8.058084865797774,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 33850
    },
    {
      "epoch": 8.058322918526454,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3551,
      "step": 33851
    },
    {
      "epoch": 8.058560971255133,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 33852
    },
    {
      "epoch": 8.058799023983813,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2622,
      "step": 33853
    },
    {
      "epoch": 8.059037076712492,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 33854
    },
    {
      "epoch": 8.059275129441172,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8909,
      "step": 33855
    },
    {
      "epoch": 8.05951318216985,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 33856
    },
    {
      "epoch": 8.05975123489853,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8367,
      "step": 33857
    },
    {
      "epoch": 8.059989287627209,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 33858
    },
    {
      "epoch": 8.060227340355889,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0428,
      "step": 33859
    },
    {
      "epoch": 8.060465393084568,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8518,
      "step": 33860
    },
    {
      "epoch": 8.060703445813248,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7746,
      "step": 33861
    },
    {
      "epoch": 8.060941498541927,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 33862
    },
    {
      "epoch": 8.061179551270607,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 33863
    },
    {
      "epoch": 8.061417603999287,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 33864
    },
    {
      "epoch": 8.061655656727964,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 33865
    },
    {
      "epoch": 8.061893709456644,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 33866
    },
    {
      "epoch": 8.062131762185324,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7766,
      "step": 33867
    },
    {
      "epoch": 8.062369814914003,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 33868
    },
    {
      "epoch": 8.062607867642683,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 33869
    },
    {
      "epoch": 8.062845920371362,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 33870
    },
    {
      "epoch": 8.063083973100042,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 33871
    },
    {
      "epoch": 8.063322025828722,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 33872
    },
    {
      "epoch": 8.063560078557401,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 33873
    },
    {
      "epoch": 8.06379813128608,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9791,
      "step": 33874
    },
    {
      "epoch": 8.064036184014759,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1399,
      "step": 33875
    },
    {
      "epoch": 8.064274236743438,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 33876
    },
    {
      "epoch": 8.064512289472118,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4292,
      "step": 33877
    },
    {
      "epoch": 8.064750342200798,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 33878
    },
    {
      "epoch": 8.064988394929477,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0672,
      "step": 33879
    },
    {
      "epoch": 8.065226447658157,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 33880
    },
    {
      "epoch": 8.065464500386836,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.6995,
      "step": 33881
    },
    {
      "epoch": 8.065702553115514,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 33882
    },
    {
      "epoch": 8.065940605844194,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 33883
    },
    {
      "epoch": 8.066178658572873,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0436,
      "step": 33884
    },
    {
      "epoch": 8.066416711301553,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 33885
    },
    {
      "epoch": 8.066654764030233,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6707,
      "step": 33886
    },
    {
      "epoch": 8.066892816758912,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 33887
    },
    {
      "epoch": 8.067130869487592,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 33888
    },
    {
      "epoch": 8.067368922216271,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 33889
    },
    {
      "epoch": 8.067606974944951,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9742,
      "step": 33890
    },
    {
      "epoch": 8.067845027673629,
      "grad_norm": 7.1875,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 33891
    },
    {
      "epoch": 8.068083080402308,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 33892
    },
    {
      "epoch": 8.068321133130988,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8825,
      "step": 33893
    },
    {
      "epoch": 8.068559185859668,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 33894
    },
    {
      "epoch": 8.068797238588347,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 33895
    },
    {
      "epoch": 8.069035291317027,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 33896
    },
    {
      "epoch": 8.069273344045707,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9247,
      "step": 33897
    },
    {
      "epoch": 8.069511396774386,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9608,
      "step": 33898
    },
    {
      "epoch": 8.069749449503066,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 33899
    },
    {
      "epoch": 8.069987502231744,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 33900
    },
    {
      "epoch": 8.069987502231744,
      "eval_runtime": 0.5049,
      "eval_samples_per_second": 1.981,
      "eval_steps_per_second": 1.981,
      "step": 33900
    },
    {
      "epoch": 8.070225554960423,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.33,
      "step": 33901
    },
    {
      "epoch": 8.070463607689103,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0621,
      "step": 33902
    },
    {
      "epoch": 8.070701660417782,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1926,
      "step": 33903
    },
    {
      "epoch": 8.070939713146462,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2323,
      "step": 33904
    },
    {
      "epoch": 8.071177765875142,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 33905
    },
    {
      "epoch": 8.071415818603821,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 33906
    },
    {
      "epoch": 8.0716538713325,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9003,
      "step": 33907
    },
    {
      "epoch": 8.07189192406118,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9605,
      "step": 33908
    },
    {
      "epoch": 8.072129976789858,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 33909
    },
    {
      "epoch": 8.072368029518538,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 33910
    },
    {
      "epoch": 8.072606082247217,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7259,
      "step": 33911
    },
    {
      "epoch": 8.072844134975897,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1885,
      "step": 33912
    },
    {
      "epoch": 8.073082187704577,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 33913
    },
    {
      "epoch": 8.073320240433256,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2367,
      "step": 33914
    },
    {
      "epoch": 8.073558293161936,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9279,
      "step": 33915
    },
    {
      "epoch": 8.073796345890615,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 33916
    },
    {
      "epoch": 8.074034398619293,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 33917
    },
    {
      "epoch": 8.074272451347973,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 33918
    },
    {
      "epoch": 8.074510504076652,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 33919
    },
    {
      "epoch": 8.074748556805332,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 33920
    },
    {
      "epoch": 8.074986609534012,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 33921
    },
    {
      "epoch": 8.075224662262691,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 33922
    },
    {
      "epoch": 8.075462714991371,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 33923
    },
    {
      "epoch": 8.07570076772005,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 33924
    },
    {
      "epoch": 8.07593882044873,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 33925
    },
    {
      "epoch": 8.076176873177408,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1416,
      "step": 33926
    },
    {
      "epoch": 8.076414925906088,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 33927
    },
    {
      "epoch": 8.076652978634767,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 33928
    },
    {
      "epoch": 8.076891031363447,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 33929
    },
    {
      "epoch": 8.077129084092126,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 33930
    },
    {
      "epoch": 8.077367136820806,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 33931
    },
    {
      "epoch": 8.077605189549486,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 33932
    },
    {
      "epoch": 8.077843242278165,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 33933
    },
    {
      "epoch": 8.078081295006845,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.4194,
      "step": 33934
    },
    {
      "epoch": 8.078319347735523,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 33935
    },
    {
      "epoch": 8.078557400464202,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1474,
      "step": 33936
    },
    {
      "epoch": 8.078795453192882,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 33937
    },
    {
      "epoch": 8.079033505921561,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 33938
    },
    {
      "epoch": 8.079271558650241,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 33939
    },
    {
      "epoch": 8.07950961137892,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.754,
      "step": 33940
    },
    {
      "epoch": 8.0797476641076,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 33941
    },
    {
      "epoch": 8.07998571683628,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 33942
    },
    {
      "epoch": 8.08022376956496,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 33943
    },
    {
      "epoch": 8.080461822293637,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 33944
    },
    {
      "epoch": 8.080699875022317,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 33945
    },
    {
      "epoch": 8.080937927750997,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 33946
    },
    {
      "epoch": 8.081175980479676,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 33947
    },
    {
      "epoch": 8.081414033208356,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.6464,
      "step": 33948
    },
    {
      "epoch": 8.081652085937035,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2064,
      "step": 33949
    },
    {
      "epoch": 8.081890138665715,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 33950
    },
    {
      "epoch": 8.082128191394395,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 33951
    },
    {
      "epoch": 8.082366244123074,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 33952
    },
    {
      "epoch": 8.082604296851752,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 33953
    },
    {
      "epoch": 8.082842349580432,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0946,
      "step": 33954
    },
    {
      "epoch": 8.083080402309111,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 33955
    },
    {
      "epoch": 8.08331845503779,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2006,
      "step": 33956
    },
    {
      "epoch": 8.08355650776647,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 33957
    },
    {
      "epoch": 8.08379456049515,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 33958
    },
    {
      "epoch": 8.08403261322383,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 33959
    },
    {
      "epoch": 8.08427066595251,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 33960
    },
    {
      "epoch": 8.084508718681187,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 33961
    },
    {
      "epoch": 8.084746771409867,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2451,
      "step": 33962
    },
    {
      "epoch": 8.084984824138546,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 33963
    },
    {
      "epoch": 8.085222876867226,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 33964
    },
    {
      "epoch": 8.085460929595905,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 33965
    },
    {
      "epoch": 8.085698982324585,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 33966
    },
    {
      "epoch": 8.085937035053265,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 33967
    },
    {
      "epoch": 8.086175087781944,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 33968
    },
    {
      "epoch": 8.086413140510624,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 33969
    },
    {
      "epoch": 8.086651193239302,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 33970
    },
    {
      "epoch": 8.086889245967981,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1936,
      "step": 33971
    },
    {
      "epoch": 8.087127298696661,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 33972
    },
    {
      "epoch": 8.08736535142534,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.173,
      "step": 33973
    },
    {
      "epoch": 8.08760340415402,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 33974
    },
    {
      "epoch": 8.0878414568827,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0718,
      "step": 33975
    },
    {
      "epoch": 8.08807950961138,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 33976
    },
    {
      "epoch": 8.088317562340059,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 33977
    },
    {
      "epoch": 8.088555615068739,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 33978
    },
    {
      "epoch": 8.088793667797416,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9967,
      "step": 33979
    },
    {
      "epoch": 8.089031720526096,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 33980
    },
    {
      "epoch": 8.089269773254776,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 33981
    },
    {
      "epoch": 8.089507825983455,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1254,
      "step": 33982
    },
    {
      "epoch": 8.089745878712135,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 33983
    },
    {
      "epoch": 8.089983931440814,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 33984
    },
    {
      "epoch": 8.090221984169494,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9528,
      "step": 33985
    },
    {
      "epoch": 8.090460036898174,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 33986
    },
    {
      "epoch": 8.090698089626853,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 33987
    },
    {
      "epoch": 8.090936142355531,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 33988
    },
    {
      "epoch": 8.09117419508421,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.027,
      "step": 33989
    },
    {
      "epoch": 8.09141224781289,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 33990
    },
    {
      "epoch": 8.09165030054157,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 33991
    },
    {
      "epoch": 8.09188835327025,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 33992
    },
    {
      "epoch": 8.092126405998929,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 33993
    },
    {
      "epoch": 8.092364458727609,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 33994
    },
    {
      "epoch": 8.092602511456288,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 33995
    },
    {
      "epoch": 8.092840564184966,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.7612,
      "step": 33996
    },
    {
      "epoch": 8.093078616913646,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 33997
    },
    {
      "epoch": 8.093316669642325,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 33998
    },
    {
      "epoch": 8.093554722371005,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9857,
      "step": 33999
    },
    {
      "epoch": 8.093792775099685,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 34000
    },
    {
      "epoch": 8.093792775099685,
      "eval_runtime": 0.5128,
      "eval_samples_per_second": 1.95,
      "eval_steps_per_second": 1.95,
      "step": 34000
    },
    {
      "epoch": 8.094030827828364,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 34001
    },
    {
      "epoch": 8.094268880557044,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 34002
    },
    {
      "epoch": 8.094506933285723,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.858,
      "step": 34003
    },
    {
      "epoch": 8.094744986014403,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1422,
      "step": 34004
    },
    {
      "epoch": 8.09498303874308,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.2785,
      "step": 34005
    },
    {
      "epoch": 8.09522109147176,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 34006
    },
    {
      "epoch": 8.09545914420044,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 34007
    },
    {
      "epoch": 8.09569719692912,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0268,
      "step": 34008
    },
    {
      "epoch": 8.0959352496578,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 34009
    },
    {
      "epoch": 8.096173302386479,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7529,
      "step": 34010
    },
    {
      "epoch": 8.096411355115158,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 34011
    },
    {
      "epoch": 8.096649407843838,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 34012
    },
    {
      "epoch": 8.096887460572518,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 34013
    },
    {
      "epoch": 8.097125513301195,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 34014
    },
    {
      "epoch": 8.097363566029875,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 34015
    },
    {
      "epoch": 8.097601618758555,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 34016
    },
    {
      "epoch": 8.097839671487234,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 34017
    },
    {
      "epoch": 8.098077724215914,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.4739,
      "step": 34018
    },
    {
      "epoch": 8.098315776944593,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7732,
      "step": 34019
    },
    {
      "epoch": 8.098553829673273,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 34020
    },
    {
      "epoch": 8.098791882401953,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9278,
      "step": 34021
    },
    {
      "epoch": 8.099029935130632,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 34022
    },
    {
      "epoch": 8.09926798785931,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 34023
    },
    {
      "epoch": 8.09950604058799,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 34024
    },
    {
      "epoch": 8.09974409331667,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 34025
    },
    {
      "epoch": 8.099982146045349,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 34026
    },
    {
      "epoch": 8.100220198774029,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 34027
    },
    {
      "epoch": 8.100458251502708,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0392,
      "step": 34028
    },
    {
      "epoch": 8.100696304231388,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 34029
    },
    {
      "epoch": 8.100934356960067,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 34030
    },
    {
      "epoch": 8.101172409688745,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 34031
    },
    {
      "epoch": 8.101410462417425,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 34032
    },
    {
      "epoch": 8.101648515146104,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 34033
    },
    {
      "epoch": 8.101886567874784,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 34034
    },
    {
      "epoch": 8.102124620603464,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2957,
      "step": 34035
    },
    {
      "epoch": 8.102362673332143,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 34036
    },
    {
      "epoch": 8.102600726060823,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 34037
    },
    {
      "epoch": 8.102838778789502,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3401,
      "step": 34038
    },
    {
      "epoch": 8.103076831518182,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 34039
    },
    {
      "epoch": 8.10331488424686,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 34040
    },
    {
      "epoch": 8.10355293697554,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.4344,
      "step": 34041
    },
    {
      "epoch": 8.103790989704219,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8124,
      "step": 34042
    },
    {
      "epoch": 8.104029042432899,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7972,
      "step": 34043
    },
    {
      "epoch": 8.104267095161578,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 34044
    },
    {
      "epoch": 8.104505147890258,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 34045
    },
    {
      "epoch": 8.104743200618937,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 34046
    },
    {
      "epoch": 8.104981253347617,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 34047
    },
    {
      "epoch": 8.105219306076297,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8798,
      "step": 34048
    },
    {
      "epoch": 8.105457358804975,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 34049
    },
    {
      "epoch": 8.105695411533654,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8007,
      "step": 34050
    },
    {
      "epoch": 8.105933464262334,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 34051
    },
    {
      "epoch": 8.106171516991013,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 34052
    },
    {
      "epoch": 8.106409569719693,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 34053
    },
    {
      "epoch": 8.106647622448373,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 34054
    },
    {
      "epoch": 8.106885675177052,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 34055
    },
    {
      "epoch": 8.107123727905732,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 34056
    },
    {
      "epoch": 8.107361780634411,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 34057
    },
    {
      "epoch": 8.10759983336309,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 34058
    },
    {
      "epoch": 8.107837886091769,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 34059
    },
    {
      "epoch": 8.108075938820448,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 34060
    },
    {
      "epoch": 8.108313991549128,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 34061
    },
    {
      "epoch": 8.108552044277808,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 34062
    },
    {
      "epoch": 8.108790097006487,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9149,
      "step": 34063
    },
    {
      "epoch": 8.109028149735167,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 34064
    },
    {
      "epoch": 8.109266202463846,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 34065
    },
    {
      "epoch": 8.109504255192524,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3181,
      "step": 34066
    },
    {
      "epoch": 8.109742307921204,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 34067
    },
    {
      "epoch": 8.109980360649883,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 34068
    },
    {
      "epoch": 8.110218413378563,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 34069
    },
    {
      "epoch": 8.110456466107243,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2752,
      "step": 34070
    },
    {
      "epoch": 8.110694518835922,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 34071
    },
    {
      "epoch": 8.110932571564602,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 34072
    },
    {
      "epoch": 8.111170624293282,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 34073
    },
    {
      "epoch": 8.111408677021961,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 34074
    },
    {
      "epoch": 8.111646729750639,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 34075
    },
    {
      "epoch": 8.111884782479319,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 34076
    },
    {
      "epoch": 8.112122835207998,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 34077
    },
    {
      "epoch": 8.112360887936678,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9685,
      "step": 34078
    },
    {
      "epoch": 8.112598940665357,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 34079
    },
    {
      "epoch": 8.112836993394037,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.8133,
      "step": 34080
    },
    {
      "epoch": 8.113075046122717,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 34081
    },
    {
      "epoch": 8.113313098851396,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 34082
    },
    {
      "epoch": 8.113551151580076,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 34083
    },
    {
      "epoch": 8.113789204308754,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9625,
      "step": 34084
    },
    {
      "epoch": 8.114027257037433,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 34085
    },
    {
      "epoch": 8.114265309766113,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 34086
    },
    {
      "epoch": 8.114503362494792,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 34087
    },
    {
      "epoch": 8.114741415223472,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2281,
      "step": 34088
    },
    {
      "epoch": 8.114979467952152,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3709,
      "step": 34089
    },
    {
      "epoch": 8.115217520680831,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.0488,
      "step": 34090
    },
    {
      "epoch": 8.11545557340951,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 34091
    },
    {
      "epoch": 8.11569362613819,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 34092
    },
    {
      "epoch": 8.115931678866868,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 34093
    },
    {
      "epoch": 8.116169731595548,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 34094
    },
    {
      "epoch": 8.116407784324227,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1829,
      "step": 34095
    },
    {
      "epoch": 8.116645837052907,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 34096
    },
    {
      "epoch": 8.116883889781587,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8781,
      "step": 34097
    },
    {
      "epoch": 8.117121942510266,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 34098
    },
    {
      "epoch": 8.117359995238946,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9234,
      "step": 34099
    },
    {
      "epoch": 8.117598047967626,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 34100
    },
    {
      "epoch": 8.117598047967626,
      "eval_runtime": 0.521,
      "eval_samples_per_second": 1.919,
      "eval_steps_per_second": 1.919,
      "step": 34100
    },
    {
      "epoch": 8.117836100696305,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.505,
      "step": 34101
    },
    {
      "epoch": 8.118074153424983,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 34102
    },
    {
      "epoch": 8.118312206153663,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 34103
    },
    {
      "epoch": 8.118550258882342,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0777,
      "step": 34104
    },
    {
      "epoch": 8.118788311611022,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1307,
      "step": 34105
    },
    {
      "epoch": 8.119026364339701,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 34106
    },
    {
      "epoch": 8.119264417068381,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 34107
    },
    {
      "epoch": 8.11950246979706,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 34108
    },
    {
      "epoch": 8.11974052252574,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1021,
      "step": 34109
    },
    {
      "epoch": 8.119978575254418,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 34110
    },
    {
      "epoch": 8.120216627983098,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 34111
    },
    {
      "epoch": 8.120454680711777,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8035,
      "step": 34112
    },
    {
      "epoch": 8.120692733440457,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 34113
    },
    {
      "epoch": 8.120930786169136,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1758,
      "step": 34114
    },
    {
      "epoch": 8.121168838897816,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8021,
      "step": 34115
    },
    {
      "epoch": 8.121406891626496,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 34116
    },
    {
      "epoch": 8.121644944355175,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 34117
    },
    {
      "epoch": 8.121882997083855,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3274,
      "step": 34118
    },
    {
      "epoch": 8.122121049812533,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2638,
      "step": 34119
    },
    {
      "epoch": 8.122359102541212,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 34120
    },
    {
      "epoch": 8.122597155269892,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1285,
      "step": 34121
    },
    {
      "epoch": 8.122835207998572,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 34122
    },
    {
      "epoch": 8.123073260727251,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 34123
    },
    {
      "epoch": 8.12331131345593,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 34124
    },
    {
      "epoch": 8.12354936618461,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 34125
    },
    {
      "epoch": 8.12378741891329,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 34126
    },
    {
      "epoch": 8.12402547164197,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.857,
      "step": 34127
    },
    {
      "epoch": 8.124263524370647,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 34128
    },
    {
      "epoch": 8.124501577099327,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 34129
    },
    {
      "epoch": 8.124739629828007,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 34130
    },
    {
      "epoch": 8.124977682556686,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2964,
      "step": 34131
    },
    {
      "epoch": 8.125215735285366,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 34132
    },
    {
      "epoch": 8.125453788014045,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 34133
    },
    {
      "epoch": 8.125691840742725,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0917,
      "step": 34134
    },
    {
      "epoch": 8.125929893471405,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 34135
    },
    {
      "epoch": 8.126167946200084,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 34136
    },
    {
      "epoch": 8.126405998928762,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 34137
    },
    {
      "epoch": 8.126644051657442,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 34138
    },
    {
      "epoch": 8.126882104386121,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 34139
    },
    {
      "epoch": 8.1271201571148,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2087,
      "step": 34140
    },
    {
      "epoch": 8.12735820984348,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 34141
    },
    {
      "epoch": 8.12759626257216,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 34142
    },
    {
      "epoch": 8.12783431530084,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 34143
    },
    {
      "epoch": 8.12807236802952,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9103,
      "step": 34144
    },
    {
      "epoch": 8.128310420758197,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9379,
      "step": 34145
    },
    {
      "epoch": 8.128548473486877,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 34146
    },
    {
      "epoch": 8.128786526215556,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0947,
      "step": 34147
    },
    {
      "epoch": 8.129024578944236,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 34148
    },
    {
      "epoch": 8.129262631672916,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 34149
    },
    {
      "epoch": 8.129500684401595,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 34150
    },
    {
      "epoch": 8.129738737130275,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 34151
    },
    {
      "epoch": 8.129976789858954,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1482,
      "step": 34152
    },
    {
      "epoch": 8.130214842587634,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 34153
    },
    {
      "epoch": 8.130452895316312,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 34154
    },
    {
      "epoch": 8.130690948044991,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3471,
      "step": 34155
    },
    {
      "epoch": 8.130929000773671,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0379,
      "step": 34156
    },
    {
      "epoch": 8.13116705350235,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 34157
    },
    {
      "epoch": 8.13140510623103,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 34158
    },
    {
      "epoch": 8.13164315895971,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 34159
    },
    {
      "epoch": 8.13188121168839,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.846,
      "step": 34160
    },
    {
      "epoch": 8.132119264417069,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 34161
    },
    {
      "epoch": 8.132357317145749,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 34162
    },
    {
      "epoch": 8.132595369874426,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 34163
    },
    {
      "epoch": 8.132833422603106,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2808,
      "step": 34164
    },
    {
      "epoch": 8.133071475331786,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 34165
    },
    {
      "epoch": 8.133309528060465,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2191,
      "step": 34166
    },
    {
      "epoch": 8.133547580789145,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 34167
    },
    {
      "epoch": 8.133785633517824,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 34168
    },
    {
      "epoch": 8.134023686246504,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 34169
    },
    {
      "epoch": 8.134261738975184,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 34170
    },
    {
      "epoch": 8.134499791703863,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 34171
    },
    {
      "epoch": 8.134737844432541,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 34172
    },
    {
      "epoch": 8.13497589716122,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.2555,
      "step": 34173
    },
    {
      "epoch": 8.1352139498899,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6741,
      "step": 34174
    },
    {
      "epoch": 8.13545200261858,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 34175
    },
    {
      "epoch": 8.13569005534726,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 34176
    },
    {
      "epoch": 8.13592810807594,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 34177
    },
    {
      "epoch": 8.136166160804619,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6649,
      "step": 34178
    },
    {
      "epoch": 8.136404213533298,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8624,
      "step": 34179
    },
    {
      "epoch": 8.136642266261976,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8887,
      "step": 34180
    },
    {
      "epoch": 8.136880318990656,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 34181
    },
    {
      "epoch": 8.137118371719335,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 34182
    },
    {
      "epoch": 8.137356424448015,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 34183
    },
    {
      "epoch": 8.137594477176695,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1884,
      "step": 34184
    },
    {
      "epoch": 8.137832529905374,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 34185
    },
    {
      "epoch": 8.138070582634054,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 34186
    },
    {
      "epoch": 8.138308635362733,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 34187
    },
    {
      "epoch": 8.138546688091413,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8613,
      "step": 34188
    },
    {
      "epoch": 8.13878474082009,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 34189
    },
    {
      "epoch": 8.13902279354877,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 34190
    },
    {
      "epoch": 8.13926084627745,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 34191
    },
    {
      "epoch": 8.13949889900613,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9515,
      "step": 34192
    },
    {
      "epoch": 8.13973695173481,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 34193
    },
    {
      "epoch": 8.139975004463489,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 34194
    },
    {
      "epoch": 8.140213057192168,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 34195
    },
    {
      "epoch": 8.140451109920848,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 34196
    },
    {
      "epoch": 8.140689162649528,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 34197
    },
    {
      "epoch": 8.140927215378206,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8055,
      "step": 34198
    },
    {
      "epoch": 8.141165268106885,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0405,
      "step": 34199
    },
    {
      "epoch": 8.141403320835565,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8024,
      "step": 34200
    },
    {
      "epoch": 8.141403320835565,
      "eval_runtime": 0.5146,
      "eval_samples_per_second": 1.943,
      "eval_steps_per_second": 1.943,
      "step": 34200
    },
    {
      "epoch": 8.141641373564244,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 34201
    },
    {
      "epoch": 8.141879426292924,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 34202
    },
    {
      "epoch": 8.142117479021604,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 34203
    },
    {
      "epoch": 8.142355531750283,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 34204
    },
    {
      "epoch": 8.142593584478963,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2228,
      "step": 34205
    },
    {
      "epoch": 8.142831637207642,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 34206
    },
    {
      "epoch": 8.14306968993632,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1577,
      "step": 34207
    },
    {
      "epoch": 8.143307742665,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 34208
    },
    {
      "epoch": 8.14354579539368,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 34209
    },
    {
      "epoch": 8.143783848122359,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 34210
    },
    {
      "epoch": 8.144021900851039,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 34211
    },
    {
      "epoch": 8.144259953579718,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 34212
    },
    {
      "epoch": 8.144498006308398,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 34213
    },
    {
      "epoch": 8.144736059037077,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 34214
    },
    {
      "epoch": 8.144974111765755,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7469,
      "step": 34215
    },
    {
      "epoch": 8.145212164494435,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7838,
      "step": 34216
    },
    {
      "epoch": 8.145450217223114,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 34217
    },
    {
      "epoch": 8.145688269951794,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9207,
      "step": 34218
    },
    {
      "epoch": 8.145926322680474,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3487,
      "step": 34219
    },
    {
      "epoch": 8.146164375409153,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.061,
      "step": 34220
    },
    {
      "epoch": 8.146402428137833,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1932,
      "step": 34221
    },
    {
      "epoch": 8.146640480866512,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 34222
    },
    {
      "epoch": 8.146878533595192,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8371,
      "step": 34223
    },
    {
      "epoch": 8.14711658632387,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 34224
    },
    {
      "epoch": 8.14735463905255,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 34225
    },
    {
      "epoch": 8.14759269178123,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7618,
      "step": 34226
    },
    {
      "epoch": 8.147830744509909,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 34227
    },
    {
      "epoch": 8.148068797238588,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7806,
      "step": 34228
    },
    {
      "epoch": 8.148306849967268,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2536,
      "step": 34229
    },
    {
      "epoch": 8.148544902695948,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 34230
    },
    {
      "epoch": 8.148782955424627,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 34231
    },
    {
      "epoch": 8.149021008153307,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2409,
      "step": 34232
    },
    {
      "epoch": 8.149259060881985,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 34233
    },
    {
      "epoch": 8.149497113610664,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 34234
    },
    {
      "epoch": 8.149735166339344,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 34235
    },
    {
      "epoch": 8.149973219068023,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 34236
    },
    {
      "epoch": 8.150211271796703,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8542,
      "step": 34237
    },
    {
      "epoch": 8.150449324525383,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8764,
      "step": 34238
    },
    {
      "epoch": 8.150687377254062,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 34239
    },
    {
      "epoch": 8.150925429982742,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 34240
    },
    {
      "epoch": 8.151163482711421,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 34241
    },
    {
      "epoch": 8.1514015354401,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 34242
    },
    {
      "epoch": 8.151639588168779,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 34243
    },
    {
      "epoch": 8.151877640897458,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 34244
    },
    {
      "epoch": 8.152115693626138,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 34245
    },
    {
      "epoch": 8.152353746354818,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1795,
      "step": 34246
    },
    {
      "epoch": 8.152591799083497,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9587,
      "step": 34247
    },
    {
      "epoch": 8.152829851812177,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 34248
    },
    {
      "epoch": 8.153067904540857,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.3625,
      "step": 34249
    },
    {
      "epoch": 8.153305957269534,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9138,
      "step": 34250
    },
    {
      "epoch": 8.153544009998214,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 34251
    },
    {
      "epoch": 8.153782062726894,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0174,
      "step": 34252
    },
    {
      "epoch": 8.154020115455573,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 34253
    },
    {
      "epoch": 8.154258168184253,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 34254
    },
    {
      "epoch": 8.154496220912932,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.946,
      "step": 34255
    },
    {
      "epoch": 8.154734273641612,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 34256
    },
    {
      "epoch": 8.154972326370292,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9922,
      "step": 34257
    },
    {
      "epoch": 8.155210379098971,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 34258
    },
    {
      "epoch": 8.155448431827649,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8299,
      "step": 34259
    },
    {
      "epoch": 8.155686484556329,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 34260
    },
    {
      "epoch": 8.155924537285008,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 34261
    },
    {
      "epoch": 8.156162590013688,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 34262
    },
    {
      "epoch": 8.156400642742367,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8847,
      "step": 34263
    },
    {
      "epoch": 8.156638695471047,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 34264
    },
    {
      "epoch": 8.156876748199727,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7945,
      "step": 34265
    },
    {
      "epoch": 8.157114800928406,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2174,
      "step": 34266
    },
    {
      "epoch": 8.157352853657086,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 34267
    },
    {
      "epoch": 8.157590906385764,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 34268
    },
    {
      "epoch": 8.157828959114443,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 34269
    },
    {
      "epoch": 8.158067011843123,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 34270
    },
    {
      "epoch": 8.158305064571802,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 34271
    },
    {
      "epoch": 8.158543117300482,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 34272
    },
    {
      "epoch": 8.158781170029162,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 34273
    },
    {
      "epoch": 8.159019222757841,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8359,
      "step": 34274
    },
    {
      "epoch": 8.159257275486521,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.006,
      "step": 34275
    },
    {
      "epoch": 8.1594953282152,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 34276
    },
    {
      "epoch": 8.159733380943878,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 34277
    },
    {
      "epoch": 8.159971433672558,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 34278
    },
    {
      "epoch": 8.160209486401238,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 34279
    },
    {
      "epoch": 8.160447539129917,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 34280
    },
    {
      "epoch": 8.160685591858597,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.3202,
      "step": 34281
    },
    {
      "epoch": 8.160923644587276,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.4705,
      "step": 34282
    },
    {
      "epoch": 8.161161697315956,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 34283
    },
    {
      "epoch": 8.161399750044636,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8157,
      "step": 34284
    },
    {
      "epoch": 8.161637802773313,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 34285
    },
    {
      "epoch": 8.161875855501993,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 34286
    },
    {
      "epoch": 8.162113908230673,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 34287
    },
    {
      "epoch": 8.162351960959352,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 34288
    },
    {
      "epoch": 8.162590013688032,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 34289
    },
    {
      "epoch": 8.162828066416711,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 34290
    },
    {
      "epoch": 8.163066119145391,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 34291
    },
    {
      "epoch": 8.16330417187407,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9026,
      "step": 34292
    },
    {
      "epoch": 8.16354222460275,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 34293
    },
    {
      "epoch": 8.163780277331428,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 34294
    },
    {
      "epoch": 8.164018330060108,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 34295
    },
    {
      "epoch": 8.164256382788787,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2767,
      "step": 34296
    },
    {
      "epoch": 8.164494435517467,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 34297
    },
    {
      "epoch": 8.164732488246147,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0042,
      "step": 34298
    },
    {
      "epoch": 8.164970540974826,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0224,
      "step": 34299
    },
    {
      "epoch": 8.165208593703506,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 34300
    },
    {
      "epoch": 8.165208593703506,
      "eval_runtime": 0.5108,
      "eval_samples_per_second": 1.958,
      "eval_steps_per_second": 1.958,
      "step": 34300
    },
    {
      "epoch": 8.165446646432185,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0636,
      "step": 34301
    },
    {
      "epoch": 8.165684699160865,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9033,
      "step": 34302
    },
    {
      "epoch": 8.165922751889543,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 34303
    },
    {
      "epoch": 8.166160804618222,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1968,
      "step": 34304
    },
    {
      "epoch": 8.166398857346902,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.279,
      "step": 34305
    },
    {
      "epoch": 8.166636910075582,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 34306
    },
    {
      "epoch": 8.166874962804261,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0378,
      "step": 34307
    },
    {
      "epoch": 8.16711301553294,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 34308
    },
    {
      "epoch": 8.16735106826162,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0778,
      "step": 34309
    },
    {
      "epoch": 8.1675891209903,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 34310
    },
    {
      "epoch": 8.16782717371898,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9214,
      "step": 34311
    },
    {
      "epoch": 8.168065226447657,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 34312
    },
    {
      "epoch": 8.168303279176337,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2423,
      "step": 34313
    },
    {
      "epoch": 8.168541331905017,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 34314
    },
    {
      "epoch": 8.168779384633696,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 34315
    },
    {
      "epoch": 8.169017437362376,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 34316
    },
    {
      "epoch": 8.169255490091055,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 34317
    },
    {
      "epoch": 8.169493542819735,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 34318
    },
    {
      "epoch": 8.169731595548415,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3666,
      "step": 34319
    },
    {
      "epoch": 8.169969648277092,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 34320
    },
    {
      "epoch": 8.170207701005772,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 34321
    },
    {
      "epoch": 8.170445753734452,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9956,
      "step": 34322
    },
    {
      "epoch": 8.170683806463131,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 34323
    },
    {
      "epoch": 8.170921859191811,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 34324
    },
    {
      "epoch": 8.17115991192049,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 34325
    },
    {
      "epoch": 8.17139796464917,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 34326
    },
    {
      "epoch": 8.17163601737785,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0343,
      "step": 34327
    },
    {
      "epoch": 8.17187407010653,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 34328
    },
    {
      "epoch": 8.172112122835207,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 34329
    },
    {
      "epoch": 8.172350175563887,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 34330
    },
    {
      "epoch": 8.172588228292566,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 34331
    },
    {
      "epoch": 8.172826281021246,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3674,
      "step": 34332
    },
    {
      "epoch": 8.173064333749926,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 34333
    },
    {
      "epoch": 8.173302386478605,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.306,
      "step": 34334
    },
    {
      "epoch": 8.173540439207285,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 34335
    },
    {
      "epoch": 8.173778491935964,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 34336
    },
    {
      "epoch": 8.174016544664644,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 34337
    },
    {
      "epoch": 8.174254597393322,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 34338
    },
    {
      "epoch": 8.174492650122001,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 34339
    },
    {
      "epoch": 8.174730702850681,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 34340
    },
    {
      "epoch": 8.17496875557936,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3755,
      "step": 34341
    },
    {
      "epoch": 8.17520680830804,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8863,
      "step": 34342
    },
    {
      "epoch": 8.17544486103672,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 34343
    },
    {
      "epoch": 8.1756829137654,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9111,
      "step": 34344
    },
    {
      "epoch": 8.175920966494079,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1994,
      "step": 34345
    },
    {
      "epoch": 8.176159019222759,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 34346
    },
    {
      "epoch": 8.176397071951437,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 34347
    },
    {
      "epoch": 8.176635124680116,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 34348
    },
    {
      "epoch": 8.176873177408796,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3688,
      "step": 34349
    },
    {
      "epoch": 8.177111230137475,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 34350
    },
    {
      "epoch": 8.177349282866155,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0113,
      "step": 34351
    },
    {
      "epoch": 8.177587335594835,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 34352
    },
    {
      "epoch": 8.177825388323514,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3003,
      "step": 34353
    },
    {
      "epoch": 8.178063441052194,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 34354
    },
    {
      "epoch": 8.178301493780872,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.213,
      "step": 34355
    },
    {
      "epoch": 8.178539546509551,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3444,
      "step": 34356
    },
    {
      "epoch": 8.17877759923823,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9352,
      "step": 34357
    },
    {
      "epoch": 8.17901565196691,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 34358
    },
    {
      "epoch": 8.17925370469559,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1206,
      "step": 34359
    },
    {
      "epoch": 8.17949175742427,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 34360
    },
    {
      "epoch": 8.17972981015295,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3135,
      "step": 34361
    },
    {
      "epoch": 8.179967862881629,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 34362
    },
    {
      "epoch": 8.180205915610308,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 34363
    },
    {
      "epoch": 8.180443968338986,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 34364
    },
    {
      "epoch": 8.180682021067666,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 34365
    },
    {
      "epoch": 8.180920073796345,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.627,
      "step": 34366
    },
    {
      "epoch": 8.181158126525025,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8782,
      "step": 34367
    },
    {
      "epoch": 8.181396179253705,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.4838,
      "step": 34368
    },
    {
      "epoch": 8.181634231982384,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 34369
    },
    {
      "epoch": 8.181872284711064,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 34370
    },
    {
      "epoch": 8.182110337439743,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 34371
    },
    {
      "epoch": 8.182348390168423,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 34372
    },
    {
      "epoch": 8.182586442897101,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 34373
    },
    {
      "epoch": 8.18282449562578,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 34374
    },
    {
      "epoch": 8.18306254835446,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 34375
    },
    {
      "epoch": 8.18330060108314,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 34376
    },
    {
      "epoch": 8.18353865381182,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9906,
      "step": 34377
    },
    {
      "epoch": 8.183776706540499,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4098,
      "step": 34378
    },
    {
      "epoch": 8.184014759269179,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 34379
    },
    {
      "epoch": 8.184252811997858,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.5134,
      "step": 34380
    },
    {
      "epoch": 8.184490864726538,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 34381
    },
    {
      "epoch": 8.184728917455216,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 34382
    },
    {
      "epoch": 8.184966970183895,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0324,
      "step": 34383
    },
    {
      "epoch": 8.185205022912575,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9409,
      "step": 34384
    },
    {
      "epoch": 8.185443075641254,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1039,
      "step": 34385
    },
    {
      "epoch": 8.185681128369934,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 34386
    },
    {
      "epoch": 8.185919181098614,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2088,
      "step": 34387
    },
    {
      "epoch": 8.186157233827293,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0711,
      "step": 34388
    },
    {
      "epoch": 8.186395286555973,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 34389
    },
    {
      "epoch": 8.186633339284652,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 34390
    },
    {
      "epoch": 8.18687139201333,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 34391
    },
    {
      "epoch": 8.18710944474201,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 34392
    },
    {
      "epoch": 8.18734749747069,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 34393
    },
    {
      "epoch": 8.187585550199369,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 34394
    },
    {
      "epoch": 8.187823602928049,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 34395
    },
    {
      "epoch": 8.188061655656728,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 34396
    },
    {
      "epoch": 8.188299708385408,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 34397
    },
    {
      "epoch": 8.188537761114087,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 34398
    },
    {
      "epoch": 8.188775813842765,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 34399
    },
    {
      "epoch": 8.189013866571445,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 34400
    },
    {
      "epoch": 8.189013866571445,
      "eval_runtime": 0.5159,
      "eval_samples_per_second": 1.938,
      "eval_steps_per_second": 1.938,
      "step": 34400
    },
    {
      "epoch": 8.189251919300125,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.4732,
      "step": 34401
    },
    {
      "epoch": 8.189489972028804,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 34402
    },
    {
      "epoch": 8.189728024757484,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 34403
    },
    {
      "epoch": 8.189966077486163,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9088,
      "step": 34404
    },
    {
      "epoch": 8.190204130214843,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 34405
    },
    {
      "epoch": 8.190442182943523,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7756,
      "step": 34406
    },
    {
      "epoch": 8.190680235672202,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7279,
      "step": 34407
    },
    {
      "epoch": 8.19091828840088,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 34408
    },
    {
      "epoch": 8.19115634112956,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8301,
      "step": 34409
    },
    {
      "epoch": 8.19139439385824,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2371,
      "step": 34410
    },
    {
      "epoch": 8.191632446586919,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8921,
      "step": 34411
    },
    {
      "epoch": 8.191870499315598,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1839,
      "step": 34412
    },
    {
      "epoch": 8.192108552044278,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 34413
    },
    {
      "epoch": 8.192346604772958,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 34414
    },
    {
      "epoch": 8.192584657501637,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 34415
    },
    {
      "epoch": 8.192822710230317,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 34416
    },
    {
      "epoch": 8.193060762958995,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 34417
    },
    {
      "epoch": 8.193298815687674,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 34418
    },
    {
      "epoch": 8.193536868416354,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 34419
    },
    {
      "epoch": 8.193774921145033,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 34420
    },
    {
      "epoch": 8.194012973873713,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9066,
      "step": 34421
    },
    {
      "epoch": 8.194251026602393,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 34422
    },
    {
      "epoch": 8.194489079331072,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9228,
      "step": 34423
    },
    {
      "epoch": 8.194727132059752,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 34424
    },
    {
      "epoch": 8.194965184788432,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 34425
    },
    {
      "epoch": 8.19520323751711,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1946,
      "step": 34426
    },
    {
      "epoch": 8.195441290245789,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 34427
    },
    {
      "epoch": 8.195679342974469,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 34428
    },
    {
      "epoch": 8.195917395703148,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 34429
    },
    {
      "epoch": 8.196155448431828,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 34430
    },
    {
      "epoch": 8.196393501160507,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 34431
    },
    {
      "epoch": 8.196631553889187,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 34432
    },
    {
      "epoch": 8.196869606617867,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0101,
      "step": 34433
    },
    {
      "epoch": 8.197107659346544,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 34434
    },
    {
      "epoch": 8.197345712075224,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.4737,
      "step": 34435
    },
    {
      "epoch": 8.197583764803904,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3028,
      "step": 34436
    },
    {
      "epoch": 8.197821817532583,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.4017,
      "step": 34437
    },
    {
      "epoch": 8.198059870261263,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 34438
    },
    {
      "epoch": 8.198297922989942,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 34439
    },
    {
      "epoch": 8.198535975718622,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 34440
    },
    {
      "epoch": 8.198774028447302,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 34441
    },
    {
      "epoch": 8.199012081175981,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 34442
    },
    {
      "epoch": 8.199250133904659,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 34443
    },
    {
      "epoch": 8.199488186633339,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.6105,
      "step": 34444
    },
    {
      "epoch": 8.199726239362018,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1137,
      "step": 34445
    },
    {
      "epoch": 8.199964292090698,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1909,
      "step": 34446
    },
    {
      "epoch": 8.200202344819377,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.879,
      "step": 34447
    },
    {
      "epoch": 8.200440397548057,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 34448
    },
    {
      "epoch": 8.200678450276737,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 34449
    },
    {
      "epoch": 8.200916503005416,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0257,
      "step": 34450
    },
    {
      "epoch": 8.201154555734096,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 34451
    },
    {
      "epoch": 8.201392608462774,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7829,
      "step": 34452
    },
    {
      "epoch": 8.201630661191453,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9213,
      "step": 34453
    },
    {
      "epoch": 8.201868713920133,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 34454
    },
    {
      "epoch": 8.202106766648813,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 34455
    },
    {
      "epoch": 8.202344819377492,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 34456
    },
    {
      "epoch": 8.202582872106172,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 34457
    },
    {
      "epoch": 8.202820924834851,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 34458
    },
    {
      "epoch": 8.203058977563531,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 34459
    },
    {
      "epoch": 8.20329703029221,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2683,
      "step": 34460
    },
    {
      "epoch": 8.203535083020888,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 34461
    },
    {
      "epoch": 8.203773135749568,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.038,
      "step": 34462
    },
    {
      "epoch": 8.204011188478248,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 34463
    },
    {
      "epoch": 8.204249241206927,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1755,
      "step": 34464
    },
    {
      "epoch": 8.204487293935607,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1636,
      "step": 34465
    },
    {
      "epoch": 8.204725346664286,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 34466
    },
    {
      "epoch": 8.204963399392966,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 34467
    },
    {
      "epoch": 8.205201452121646,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 34468
    },
    {
      "epoch": 8.205439504850323,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1524,
      "step": 34469
    },
    {
      "epoch": 8.205677557579003,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 34470
    },
    {
      "epoch": 8.205915610307683,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9772,
      "step": 34471
    },
    {
      "epoch": 8.206153663036362,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 34472
    },
    {
      "epoch": 8.206391715765042,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 34473
    },
    {
      "epoch": 8.206629768493721,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 34474
    },
    {
      "epoch": 8.206867821222401,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 34475
    },
    {
      "epoch": 8.20710587395108,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 34476
    },
    {
      "epoch": 8.20734392667976,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7833,
      "step": 34477
    },
    {
      "epoch": 8.207581979408438,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 34478
    },
    {
      "epoch": 8.207820032137118,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 34479
    },
    {
      "epoch": 8.208058084865797,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 34480
    },
    {
      "epoch": 8.208296137594477,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 34481
    },
    {
      "epoch": 8.208534190323157,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 34482
    },
    {
      "epoch": 8.208772243051836,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 34483
    },
    {
      "epoch": 8.209010295780516,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 34484
    },
    {
      "epoch": 8.209248348509195,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 34485
    },
    {
      "epoch": 8.209486401237875,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 34486
    },
    {
      "epoch": 8.209724453966553,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 34487
    },
    {
      "epoch": 8.209962506695232,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 34488
    },
    {
      "epoch": 8.210200559423912,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 34489
    },
    {
      "epoch": 8.210438612152592,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 34490
    },
    {
      "epoch": 8.210676664881271,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0738,
      "step": 34491
    },
    {
      "epoch": 8.21091471760995,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 34492
    },
    {
      "epoch": 8.21115277033863,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9619,
      "step": 34493
    },
    {
      "epoch": 8.21139082306731,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 34494
    },
    {
      "epoch": 8.21162887579599,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 34495
    },
    {
      "epoch": 8.211866928524667,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.4107,
      "step": 34496
    },
    {
      "epoch": 8.212104981253347,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 34497
    },
    {
      "epoch": 8.212343033982027,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 34498
    },
    {
      "epoch": 8.212581086710706,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.949,
      "step": 34499
    },
    {
      "epoch": 8.212819139439386,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 34500
    },
    {
      "epoch": 8.212819139439386,
      "eval_runtime": 0.5178,
      "eval_samples_per_second": 1.931,
      "eval_steps_per_second": 1.931,
      "step": 34500
    },
    {
      "epoch": 8.213057192168066,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 34501
    },
    {
      "epoch": 8.213295244896745,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 34502
    },
    {
      "epoch": 8.213533297625425,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 34503
    },
    {
      "epoch": 8.213771350354104,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8079,
      "step": 34504
    },
    {
      "epoch": 8.214009403082782,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1201,
      "step": 34505
    },
    {
      "epoch": 8.214247455811462,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 34506
    },
    {
      "epoch": 8.214485508540141,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 34507
    },
    {
      "epoch": 8.214723561268821,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2216,
      "step": 34508
    },
    {
      "epoch": 8.2149616139975,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 34509
    },
    {
      "epoch": 8.21519966672618,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 34510
    },
    {
      "epoch": 8.21543771945486,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 34511
    },
    {
      "epoch": 8.21567577218354,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 34512
    },
    {
      "epoch": 8.215913824912217,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 34513
    },
    {
      "epoch": 8.216151877640897,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 34514
    },
    {
      "epoch": 8.216389930369576,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3312,
      "step": 34515
    },
    {
      "epoch": 8.216627983098256,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 34516
    },
    {
      "epoch": 8.216866035826936,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 34517
    },
    {
      "epoch": 8.217104088555615,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 34518
    },
    {
      "epoch": 8.217342141284295,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2041,
      "step": 34519
    },
    {
      "epoch": 8.217580194012974,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 34520
    },
    {
      "epoch": 8.217818246741654,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 34521
    },
    {
      "epoch": 8.218056299470332,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 34522
    },
    {
      "epoch": 8.218294352199011,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 34523
    },
    {
      "epoch": 8.218532404927691,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 34524
    },
    {
      "epoch": 8.21877045765637,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1403,
      "step": 34525
    },
    {
      "epoch": 8.21900851038505,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2077,
      "step": 34526
    },
    {
      "epoch": 8.21924656311373,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 34527
    },
    {
      "epoch": 8.21948461584241,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 34528
    },
    {
      "epoch": 8.21972266857109,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 34529
    },
    {
      "epoch": 8.219960721299769,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 34530
    },
    {
      "epoch": 8.220198774028447,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 34531
    },
    {
      "epoch": 8.220436826757126,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.3614,
      "step": 34532
    },
    {
      "epoch": 8.220674879485806,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9736,
      "step": 34533
    },
    {
      "epoch": 8.220912932214485,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0981,
      "step": 34534
    },
    {
      "epoch": 8.221150984943165,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 34535
    },
    {
      "epoch": 8.221389037671845,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 34536
    },
    {
      "epoch": 8.221627090400524,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 34537
    },
    {
      "epoch": 8.221865143129204,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 34538
    },
    {
      "epoch": 8.222103195857883,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 34539
    },
    {
      "epoch": 8.222341248586561,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 34540
    },
    {
      "epoch": 8.22257930131524,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 34541
    },
    {
      "epoch": 8.22281735404392,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 34542
    },
    {
      "epoch": 8.2230554067726,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 34543
    },
    {
      "epoch": 8.22329345950128,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 34544
    },
    {
      "epoch": 8.22353151222996,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 34545
    },
    {
      "epoch": 8.223769564958639,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 34546
    },
    {
      "epoch": 8.224007617687318,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 34547
    },
    {
      "epoch": 8.224245670415996,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 34548
    },
    {
      "epoch": 8.224483723144676,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.247,
      "step": 34549
    },
    {
      "epoch": 8.224721775873356,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 34550
    },
    {
      "epoch": 8.224959828602035,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 34551
    },
    {
      "epoch": 8.225197881330715,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8279,
      "step": 34552
    },
    {
      "epoch": 8.225435934059394,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 34553
    },
    {
      "epoch": 8.225673986788074,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 34554
    },
    {
      "epoch": 8.225912039516754,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 34555
    },
    {
      "epoch": 8.226150092245433,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 34556
    },
    {
      "epoch": 8.226388144974111,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2184,
      "step": 34557
    },
    {
      "epoch": 8.22662619770279,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8346,
      "step": 34558
    },
    {
      "epoch": 8.22686425043147,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 34559
    },
    {
      "epoch": 8.22710230316015,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 34560
    },
    {
      "epoch": 8.22734035588883,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 34561
    },
    {
      "epoch": 8.227578408617509,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7841,
      "step": 34562
    },
    {
      "epoch": 8.227816461346189,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8091,
      "step": 34563
    },
    {
      "epoch": 8.228054514074868,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0468,
      "step": 34564
    },
    {
      "epoch": 8.228292566803548,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 34565
    },
    {
      "epoch": 8.228530619532226,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 34566
    },
    {
      "epoch": 8.228768672260905,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 34567
    },
    {
      "epoch": 8.229006724989585,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 34568
    },
    {
      "epoch": 8.229244777718264,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 34569
    },
    {
      "epoch": 8.229482830446944,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8094,
      "step": 34570
    },
    {
      "epoch": 8.229720883175624,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 34571
    },
    {
      "epoch": 8.229958935904303,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 34572
    },
    {
      "epoch": 8.230196988632983,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 34573
    },
    {
      "epoch": 8.230435041361662,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 34574
    },
    {
      "epoch": 8.23067309409034,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 34575
    },
    {
      "epoch": 8.23091114681902,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 34576
    },
    {
      "epoch": 8.2311491995477,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 34577
    },
    {
      "epoch": 8.23138725227638,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 34578
    },
    {
      "epoch": 8.231625305005059,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1386,
      "step": 34579
    },
    {
      "epoch": 8.231863357733738,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 34580
    },
    {
      "epoch": 8.232101410462418,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 34581
    },
    {
      "epoch": 8.232339463191098,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0681,
      "step": 34582
    },
    {
      "epoch": 8.232577515919775,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 34583
    },
    {
      "epoch": 8.232815568648455,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3094,
      "step": 34584
    },
    {
      "epoch": 8.233053621377135,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 34585
    },
    {
      "epoch": 8.233291674105814,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 34586
    },
    {
      "epoch": 8.233529726834494,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 34587
    },
    {
      "epoch": 8.233767779563173,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 34588
    },
    {
      "epoch": 8.234005832291853,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 34589
    },
    {
      "epoch": 8.234243885020533,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 34590
    },
    {
      "epoch": 8.234481937749212,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 34591
    },
    {
      "epoch": 8.23471999047789,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0012,
      "step": 34592
    },
    {
      "epoch": 8.23495804320657,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3365,
      "step": 34593
    },
    {
      "epoch": 8.23519609593525,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 34594
    },
    {
      "epoch": 8.235434148663929,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9348,
      "step": 34595
    },
    {
      "epoch": 8.235672201392608,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 34596
    },
    {
      "epoch": 8.235910254121288,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 34597
    },
    {
      "epoch": 8.236148306849968,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0107,
      "step": 34598
    },
    {
      "epoch": 8.236386359578647,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0303,
      "step": 34599
    },
    {
      "epoch": 8.236624412307327,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 34600
    },
    {
      "epoch": 8.236624412307327,
      "eval_runtime": 0.5343,
      "eval_samples_per_second": 1.871,
      "eval_steps_per_second": 1.871,
      "step": 34600
    },
    {
      "epoch": 8.236862465036005,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1024,
      "step": 34601
    },
    {
      "epoch": 8.237100517764684,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 34602
    },
    {
      "epoch": 8.237338570493364,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 34603
    },
    {
      "epoch": 8.237576623222044,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 34604
    },
    {
      "epoch": 8.237814675950723,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 34605
    },
    {
      "epoch": 8.238052728679403,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 34606
    },
    {
      "epoch": 8.238290781408082,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2619,
      "step": 34607
    },
    {
      "epoch": 8.238528834136762,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1531,
      "step": 34608
    },
    {
      "epoch": 8.238766886865442,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 34609
    },
    {
      "epoch": 8.23900493959412,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 34610
    },
    {
      "epoch": 8.239242992322799,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 34611
    },
    {
      "epoch": 8.239481045051479,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 34612
    },
    {
      "epoch": 8.239719097780158,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2885,
      "step": 34613
    },
    {
      "epoch": 8.239957150508838,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 34614
    },
    {
      "epoch": 8.240195203237517,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 34615
    },
    {
      "epoch": 8.240433255966197,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7058,
      "step": 34616
    },
    {
      "epoch": 8.240671308694877,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0774,
      "step": 34617
    },
    {
      "epoch": 8.240909361423554,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 34618
    },
    {
      "epoch": 8.241147414152234,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 34619
    },
    {
      "epoch": 8.241385466880914,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7695,
      "step": 34620
    },
    {
      "epoch": 8.241623519609593,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 34621
    },
    {
      "epoch": 8.241861572338273,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 34622
    },
    {
      "epoch": 8.242099625066952,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9242,
      "step": 34623
    },
    {
      "epoch": 8.242337677795632,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 34624
    },
    {
      "epoch": 8.242575730524312,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9552,
      "step": 34625
    },
    {
      "epoch": 8.242813783252991,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 34626
    },
    {
      "epoch": 8.24305183598167,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 34627
    },
    {
      "epoch": 8.243289888710349,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 34628
    },
    {
      "epoch": 8.243527941439028,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 34629
    },
    {
      "epoch": 8.243765994167708,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 34630
    },
    {
      "epoch": 8.244004046896388,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 34631
    },
    {
      "epoch": 8.244242099625067,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9458,
      "step": 34632
    },
    {
      "epoch": 8.244480152353747,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1214,
      "step": 34633
    },
    {
      "epoch": 8.244718205082426,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8195,
      "step": 34634
    },
    {
      "epoch": 8.244956257811106,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 34635
    },
    {
      "epoch": 8.245194310539784,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2185,
      "step": 34636
    },
    {
      "epoch": 8.245432363268463,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 34637
    },
    {
      "epoch": 8.245670415997143,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 34638
    },
    {
      "epoch": 8.245908468725823,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 34639
    },
    {
      "epoch": 8.246146521454502,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 34640
    },
    {
      "epoch": 8.246384574183182,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 34641
    },
    {
      "epoch": 8.246622626911861,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9434,
      "step": 34642
    },
    {
      "epoch": 8.246860679640541,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1026,
      "step": 34643
    },
    {
      "epoch": 8.24709873236922,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1903,
      "step": 34644
    },
    {
      "epoch": 8.247336785097898,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 34645
    },
    {
      "epoch": 8.247574837826578,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 34646
    },
    {
      "epoch": 8.247812890555258,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 34647
    },
    {
      "epoch": 8.248050943283937,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 34648
    },
    {
      "epoch": 8.248288996012617,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 34649
    },
    {
      "epoch": 8.248527048741296,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 34650
    },
    {
      "epoch": 8.248765101469976,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8472,
      "step": 34651
    },
    {
      "epoch": 8.249003154198656,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8794,
      "step": 34652
    },
    {
      "epoch": 8.249241206927334,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 34653
    },
    {
      "epoch": 8.249479259656013,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 34654
    },
    {
      "epoch": 8.249717312384693,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 34655
    },
    {
      "epoch": 8.249955365113372,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 34656
    },
    {
      "epoch": 8.250193417842052,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9405,
      "step": 34657
    },
    {
      "epoch": 8.250431470570732,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2286,
      "step": 34658
    },
    {
      "epoch": 8.250669523299411,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.305,
      "step": 34659
    },
    {
      "epoch": 8.25090757602809,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 34660
    },
    {
      "epoch": 8.25114562875677,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 34661
    },
    {
      "epoch": 8.251383681485448,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 34662
    },
    {
      "epoch": 8.251621734214128,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0364,
      "step": 34663
    },
    {
      "epoch": 8.251859786942807,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8098,
      "step": 34664
    },
    {
      "epoch": 8.252097839671487,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 34665
    },
    {
      "epoch": 8.252335892400167,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 34666
    },
    {
      "epoch": 8.252573945128846,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8306,
      "step": 34667
    },
    {
      "epoch": 8.252811997857526,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3344,
      "step": 34668
    },
    {
      "epoch": 8.253050050586205,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 34669
    },
    {
      "epoch": 8.253288103314885,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 34670
    },
    {
      "epoch": 8.253526156043563,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 34671
    },
    {
      "epoch": 8.253764208772242,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 34672
    },
    {
      "epoch": 8.254002261500922,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 34673
    },
    {
      "epoch": 8.254240314229602,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 34674
    },
    {
      "epoch": 8.254478366958281,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 34675
    },
    {
      "epoch": 8.254716419686961,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 34676
    },
    {
      "epoch": 8.25495447241564,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.904,
      "step": 34677
    },
    {
      "epoch": 8.25519252514432,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 34678
    },
    {
      "epoch": 8.255430577873,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1141,
      "step": 34679
    },
    {
      "epoch": 8.255668630601678,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 34680
    },
    {
      "epoch": 8.255906683330357,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 34681
    },
    {
      "epoch": 8.256144736059037,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 34682
    },
    {
      "epoch": 8.256382788787716,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 34683
    },
    {
      "epoch": 8.256620841516396,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 34684
    },
    {
      "epoch": 8.256858894245076,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.9655,
      "step": 34685
    },
    {
      "epoch": 8.257096946973755,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 34686
    },
    {
      "epoch": 8.257334999702435,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 34687
    },
    {
      "epoch": 8.257573052431113,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 34688
    },
    {
      "epoch": 8.257811105159792,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 34689
    },
    {
      "epoch": 8.258049157888472,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 34690
    },
    {
      "epoch": 8.258287210617151,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9988,
      "step": 34691
    },
    {
      "epoch": 8.258525263345831,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 34692
    },
    {
      "epoch": 8.25876331607451,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 34693
    },
    {
      "epoch": 8.25900136880319,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.6911,
      "step": 34694
    },
    {
      "epoch": 8.25923942153187,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3538,
      "step": 34695
    },
    {
      "epoch": 8.25947747426055,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0863,
      "step": 34696
    },
    {
      "epoch": 8.259715526989227,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8487,
      "step": 34697
    },
    {
      "epoch": 8.259953579717907,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 34698
    },
    {
      "epoch": 8.260191632446586,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 34699
    },
    {
      "epoch": 8.260429685175266,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3712,
      "step": 34700
    },
    {
      "epoch": 8.260429685175266,
      "eval_runtime": 0.5263,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 34700
    },
    {
      "epoch": 8.260667737903946,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 34701
    },
    {
      "epoch": 8.260905790632625,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 34702
    },
    {
      "epoch": 8.261143843361305,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 34703
    },
    {
      "epoch": 8.261381896089985,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 34704
    },
    {
      "epoch": 8.261619948818664,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 34705
    },
    {
      "epoch": 8.261858001547342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 34706
    },
    {
      "epoch": 8.262096054276022,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9076,
      "step": 34707
    },
    {
      "epoch": 8.262334107004701,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9759,
      "step": 34708
    },
    {
      "epoch": 8.26257215973338,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 34709
    },
    {
      "epoch": 8.26281021246206,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 34710
    },
    {
      "epoch": 8.26304826519074,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 34711
    },
    {
      "epoch": 8.26328631791942,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.4414,
      "step": 34712
    },
    {
      "epoch": 8.2635243706481,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2437,
      "step": 34713
    },
    {
      "epoch": 8.263762423376779,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 34714
    },
    {
      "epoch": 8.264000476105457,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 34715
    },
    {
      "epoch": 8.264238528834136,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 34716
    },
    {
      "epoch": 8.264476581562816,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 34717
    },
    {
      "epoch": 8.264714634291495,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.037,
      "step": 34718
    },
    {
      "epoch": 8.264952687020175,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8337,
      "step": 34719
    },
    {
      "epoch": 8.265190739748855,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 34720
    },
    {
      "epoch": 8.265428792477534,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 34721
    },
    {
      "epoch": 8.265666845206214,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 34722
    },
    {
      "epoch": 8.265904897934892,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 34723
    },
    {
      "epoch": 8.266142950663571,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.86,
      "step": 34724
    },
    {
      "epoch": 8.266381003392251,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 34725
    },
    {
      "epoch": 8.26661905612093,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 34726
    },
    {
      "epoch": 8.26685710884961,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 34727
    },
    {
      "epoch": 8.26709516157829,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 34728
    },
    {
      "epoch": 8.26733321430697,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 34729
    },
    {
      "epoch": 8.267571267035649,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1411,
      "step": 34730
    },
    {
      "epoch": 8.267809319764329,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 34731
    },
    {
      "epoch": 8.268047372493006,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2392,
      "step": 34732
    },
    {
      "epoch": 8.268285425221686,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9678,
      "step": 34733
    },
    {
      "epoch": 8.268523477950366,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 34734
    },
    {
      "epoch": 8.268761530679045,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 34735
    },
    {
      "epoch": 8.268999583407725,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.163,
      "step": 34736
    },
    {
      "epoch": 8.269237636136404,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 34737
    },
    {
      "epoch": 8.269475688865084,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8563,
      "step": 34738
    },
    {
      "epoch": 8.269713741593764,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 34739
    },
    {
      "epoch": 8.269951794322443,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9199,
      "step": 34740
    },
    {
      "epoch": 8.270189847051121,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2684,
      "step": 34741
    },
    {
      "epoch": 8.2704278997798,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 34742
    },
    {
      "epoch": 8.27066595250848,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 34743
    },
    {
      "epoch": 8.27090400523716,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 34744
    },
    {
      "epoch": 8.27114205796584,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 34745
    },
    {
      "epoch": 8.271380110694519,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1736,
      "step": 34746
    },
    {
      "epoch": 8.271618163423199,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 34747
    },
    {
      "epoch": 8.271856216151878,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 34748
    },
    {
      "epoch": 8.272094268880558,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 34749
    },
    {
      "epoch": 8.272332321609236,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0212,
      "step": 34750
    },
    {
      "epoch": 8.272570374337915,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 34751
    },
    {
      "epoch": 8.272808427066595,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 34752
    },
    {
      "epoch": 8.273046479795275,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1096,
      "step": 34753
    },
    {
      "epoch": 8.273284532523954,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.6299,
      "step": 34754
    },
    {
      "epoch": 8.273522585252634,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 34755
    },
    {
      "epoch": 8.273760637981313,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 34756
    },
    {
      "epoch": 8.273998690709993,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.303,
      "step": 34757
    },
    {
      "epoch": 8.27423674343867,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 34758
    },
    {
      "epoch": 8.27447479616735,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 34759
    },
    {
      "epoch": 8.27471284889603,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 34760
    },
    {
      "epoch": 8.27495090162471,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 34761
    },
    {
      "epoch": 8.27518895435339,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 34762
    },
    {
      "epoch": 8.275427007082069,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 34763
    },
    {
      "epoch": 8.275665059810748,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 34764
    },
    {
      "epoch": 8.275903112539428,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 34765
    },
    {
      "epoch": 8.276141165268108,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 34766
    },
    {
      "epoch": 8.276379217996785,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8838,
      "step": 34767
    },
    {
      "epoch": 8.276617270725465,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2942,
      "step": 34768
    },
    {
      "epoch": 8.276855323454145,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3535,
      "step": 34769
    },
    {
      "epoch": 8.277093376182824,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3077,
      "step": 34770
    },
    {
      "epoch": 8.277331428911504,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 34771
    },
    {
      "epoch": 8.277569481640183,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9283,
      "step": 34772
    },
    {
      "epoch": 8.277807534368863,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 34773
    },
    {
      "epoch": 8.278045587097543,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 34774
    },
    {
      "epoch": 8.278283639826222,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1759,
      "step": 34775
    },
    {
      "epoch": 8.2785216925549,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 34776
    },
    {
      "epoch": 8.27875974528358,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 34777
    },
    {
      "epoch": 8.27899779801226,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 34778
    },
    {
      "epoch": 8.279235850740939,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9385,
      "step": 34779
    },
    {
      "epoch": 8.279473903469619,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 34780
    },
    {
      "epoch": 8.279711956198298,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 34781
    },
    {
      "epoch": 8.279950008926978,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 34782
    },
    {
      "epoch": 8.280188061655657,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0525,
      "step": 34783
    },
    {
      "epoch": 8.280426114384337,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 34784
    },
    {
      "epoch": 8.280664167113015,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.6836,
      "step": 34785
    },
    {
      "epoch": 8.280902219841694,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 34786
    },
    {
      "epoch": 8.281140272570374,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 34787
    },
    {
      "epoch": 8.281378325299054,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 34788
    },
    {
      "epoch": 8.281616378027733,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.071,
      "step": 34789
    },
    {
      "epoch": 8.281854430756413,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 34790
    },
    {
      "epoch": 8.282092483485092,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 34791
    },
    {
      "epoch": 8.282330536213772,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 34792
    },
    {
      "epoch": 8.28256858894245,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 34793
    },
    {
      "epoch": 8.28280664167113,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 34794
    },
    {
      "epoch": 8.283044694399809,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6124,
      "step": 34795
    },
    {
      "epoch": 8.283282747128489,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9071,
      "step": 34796
    },
    {
      "epoch": 8.283520799857168,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1067,
      "step": 34797
    },
    {
      "epoch": 8.283758852585848,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2372,
      "step": 34798
    },
    {
      "epoch": 8.283996905314527,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 34799
    },
    {
      "epoch": 8.284234958043207,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 34800
    },
    {
      "epoch": 8.284234958043207,
      "eval_runtime": 0.5214,
      "eval_samples_per_second": 1.918,
      "eval_steps_per_second": 1.918,
      "step": 34800
    },
    {
      "epoch": 8.284473010771887,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 34801
    },
    {
      "epoch": 8.284711063500565,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7216,
      "step": 34802
    },
    {
      "epoch": 8.284949116229244,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 34803
    },
    {
      "epoch": 8.285187168957924,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 34804
    },
    {
      "epoch": 8.285425221686603,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1493,
      "step": 34805
    },
    {
      "epoch": 8.285663274415283,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 34806
    },
    {
      "epoch": 8.285901327143963,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 34807
    },
    {
      "epoch": 8.286139379872642,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 34808
    },
    {
      "epoch": 8.286377432601322,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 34809
    },
    {
      "epoch": 8.286615485330001,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5122,
      "step": 34810
    },
    {
      "epoch": 8.28685353805868,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 34811
    },
    {
      "epoch": 8.287091590787359,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 34812
    },
    {
      "epoch": 8.287329643516038,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 34813
    },
    {
      "epoch": 8.287567696244718,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9868,
      "step": 34814
    },
    {
      "epoch": 8.287805748973398,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1698,
      "step": 34815
    },
    {
      "epoch": 8.288043801702077,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 34816
    },
    {
      "epoch": 8.288281854430757,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 34817
    },
    {
      "epoch": 8.288519907159436,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 34818
    },
    {
      "epoch": 8.288757959888116,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1895,
      "step": 34819
    },
    {
      "epoch": 8.288996012616794,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1603,
      "step": 34820
    },
    {
      "epoch": 8.289234065345473,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 34821
    },
    {
      "epoch": 8.289472118074153,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 34822
    },
    {
      "epoch": 8.289710170802833,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 34823
    },
    {
      "epoch": 8.289948223531512,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 34824
    },
    {
      "epoch": 8.290186276260192,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 34825
    },
    {
      "epoch": 8.290424328988871,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 34826
    },
    {
      "epoch": 8.290662381717551,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 34827
    },
    {
      "epoch": 8.290900434446229,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 34828
    },
    {
      "epoch": 8.291138487174909,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 34829
    },
    {
      "epoch": 8.291376539903588,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8486,
      "step": 34830
    },
    {
      "epoch": 8.291614592632268,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 34831
    },
    {
      "epoch": 8.291852645360947,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 34832
    },
    {
      "epoch": 8.292090698089627,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 34833
    },
    {
      "epoch": 8.292328750818307,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9739,
      "step": 34834
    },
    {
      "epoch": 8.292566803546986,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 34835
    },
    {
      "epoch": 8.292804856275666,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8117,
      "step": 34836
    },
    {
      "epoch": 8.293042909004345,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 34837
    },
    {
      "epoch": 8.293280961733023,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 34838
    },
    {
      "epoch": 8.293519014461703,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 34839
    },
    {
      "epoch": 8.293757067190382,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 34840
    },
    {
      "epoch": 8.293995119919062,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3894,
      "step": 34841
    },
    {
      "epoch": 8.294233172647742,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1378,
      "step": 34842
    },
    {
      "epoch": 8.294471225376421,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 34843
    },
    {
      "epoch": 8.2947092781051,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 34844
    },
    {
      "epoch": 8.29494733083378,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7176,
      "step": 34845
    },
    {
      "epoch": 8.295185383562458,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2576,
      "step": 34846
    },
    {
      "epoch": 8.295423436291138,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 34847
    },
    {
      "epoch": 8.295661489019817,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2807,
      "step": 34848
    },
    {
      "epoch": 8.295899541748497,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 34849
    },
    {
      "epoch": 8.296137594477177,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7257,
      "step": 34850
    },
    {
      "epoch": 8.296375647205856,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 34851
    },
    {
      "epoch": 8.296613699934536,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 34852
    },
    {
      "epoch": 8.296851752663216,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 34853
    },
    {
      "epoch": 8.297089805391895,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 34854
    },
    {
      "epoch": 8.297327858120573,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 34855
    },
    {
      "epoch": 8.297565910849253,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 34856
    },
    {
      "epoch": 8.297803963577932,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.92,
      "step": 34857
    },
    {
      "epoch": 8.298042016306612,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 34858
    },
    {
      "epoch": 8.298280069035291,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 34859
    },
    {
      "epoch": 8.298518121763971,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7127,
      "step": 34860
    },
    {
      "epoch": 8.29875617449265,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 34861
    },
    {
      "epoch": 8.29899422722133,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 34862
    },
    {
      "epoch": 8.29923227995001,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2975,
      "step": 34863
    },
    {
      "epoch": 8.299470332678688,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 34864
    },
    {
      "epoch": 8.299708385407367,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 34865
    },
    {
      "epoch": 8.299946438136047,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 34866
    },
    {
      "epoch": 8.300184490864726,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9804,
      "step": 34867
    },
    {
      "epoch": 8.300422543593406,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 34868
    },
    {
      "epoch": 8.300660596322086,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2241,
      "step": 34869
    },
    {
      "epoch": 8.300898649050765,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 34870
    },
    {
      "epoch": 8.301136701779445,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 34871
    },
    {
      "epoch": 8.301374754508124,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 34872
    },
    {
      "epoch": 8.301612807236802,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 34873
    },
    {
      "epoch": 8.301850859965482,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 34874
    },
    {
      "epoch": 8.302088912694161,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8014,
      "step": 34875
    },
    {
      "epoch": 8.302326965422841,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7591,
      "step": 34876
    },
    {
      "epoch": 8.30256501815152,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0908,
      "step": 34877
    },
    {
      "epoch": 8.3028030708802,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6021,
      "step": 34878
    },
    {
      "epoch": 8.30304112360888,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9992,
      "step": 34879
    },
    {
      "epoch": 8.30327917633756,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 34880
    },
    {
      "epoch": 8.303517229066237,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 34881
    },
    {
      "epoch": 8.303755281794917,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 34882
    },
    {
      "epoch": 8.303993334523597,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4853,
      "step": 34883
    },
    {
      "epoch": 8.304231387252276,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 34884
    },
    {
      "epoch": 8.304469439980956,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 34885
    },
    {
      "epoch": 8.304707492709635,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1578,
      "step": 34886
    },
    {
      "epoch": 8.304945545438315,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 34887
    },
    {
      "epoch": 8.305183598166995,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1568,
      "step": 34888
    },
    {
      "epoch": 8.305421650895674,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 34889
    },
    {
      "epoch": 8.305659703624352,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.5304,
      "step": 34890
    },
    {
      "epoch": 8.305897756353032,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 34891
    },
    {
      "epoch": 8.306135809081711,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 34892
    },
    {
      "epoch": 8.30637386181039,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 34893
    },
    {
      "epoch": 8.30661191453907,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.7849,
      "step": 34894
    },
    {
      "epoch": 8.30684996726775,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8131,
      "step": 34895
    },
    {
      "epoch": 8.30708801999643,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 34896
    },
    {
      "epoch": 8.30732607272511,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9148,
      "step": 34897
    },
    {
      "epoch": 8.307564125453789,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 34898
    },
    {
      "epoch": 8.307802178182467,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 34899
    },
    {
      "epoch": 8.308040230911146,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9211,
      "step": 34900
    },
    {
      "epoch": 8.308040230911146,
      "eval_runtime": 0.5226,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 34900
    },
    {
      "epoch": 8.308278283639826,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 34901
    },
    {
      "epoch": 8.308516336368506,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 34902
    },
    {
      "epoch": 8.308754389097185,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 34903
    },
    {
      "epoch": 8.308992441825865,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 34904
    },
    {
      "epoch": 8.309230494554544,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9051,
      "step": 34905
    },
    {
      "epoch": 8.309468547283224,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 34906
    },
    {
      "epoch": 8.309706600011904,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 34907
    },
    {
      "epoch": 8.309944652740581,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 34908
    },
    {
      "epoch": 8.310182705469261,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0603,
      "step": 34909
    },
    {
      "epoch": 8.31042075819794,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 34910
    },
    {
      "epoch": 8.31065881092662,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9171,
      "step": 34911
    },
    {
      "epoch": 8.3108968636553,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 34912
    },
    {
      "epoch": 8.31113491638398,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 34913
    },
    {
      "epoch": 8.311372969112659,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3261,
      "step": 34914
    },
    {
      "epoch": 8.311611021841339,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9162,
      "step": 34915
    },
    {
      "epoch": 8.311849074570016,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1573,
      "step": 34916
    },
    {
      "epoch": 8.312087127298696,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1186,
      "step": 34917
    },
    {
      "epoch": 8.312325180027376,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8402,
      "step": 34918
    },
    {
      "epoch": 8.312563232756055,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2026,
      "step": 34919
    },
    {
      "epoch": 8.312801285484735,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 34920
    },
    {
      "epoch": 8.313039338213414,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9381,
      "step": 34921
    },
    {
      "epoch": 8.313277390942094,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 34922
    },
    {
      "epoch": 8.313515443670774,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 34923
    },
    {
      "epoch": 8.313753496399453,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2406,
      "step": 34924
    },
    {
      "epoch": 8.313991549128131,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3081,
      "step": 34925
    },
    {
      "epoch": 8.31422960185681,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2527,
      "step": 34926
    },
    {
      "epoch": 8.31446765458549,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 34927
    },
    {
      "epoch": 8.31470570731417,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0104,
      "step": 34928
    },
    {
      "epoch": 8.31494376004285,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0548,
      "step": 34929
    },
    {
      "epoch": 8.31518181277153,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 34930
    },
    {
      "epoch": 8.315419865500209,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 34931
    },
    {
      "epoch": 8.315657918228888,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 34932
    },
    {
      "epoch": 8.315895970957568,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 34933
    },
    {
      "epoch": 8.316134023686246,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 34934
    },
    {
      "epoch": 8.316372076414925,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 34935
    },
    {
      "epoch": 8.316610129143605,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 34936
    },
    {
      "epoch": 8.316848181872285,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 34937
    },
    {
      "epoch": 8.317086234600964,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8198,
      "step": 34938
    },
    {
      "epoch": 8.317324287329644,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8741,
      "step": 34939
    },
    {
      "epoch": 8.317562340058323,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 34940
    },
    {
      "epoch": 8.317800392787003,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 34941
    },
    {
      "epoch": 8.318038445515683,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 34942
    },
    {
      "epoch": 8.31827649824436,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9303,
      "step": 34943
    },
    {
      "epoch": 8.31851455097304,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 34944
    },
    {
      "epoch": 8.31875260370172,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 34945
    },
    {
      "epoch": 8.3189906564304,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 34946
    },
    {
      "epoch": 8.319228709159079,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 34947
    },
    {
      "epoch": 8.319466761887758,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 34948
    },
    {
      "epoch": 8.319704814616438,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2188,
      "step": 34949
    },
    {
      "epoch": 8.319942867345118,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8112,
      "step": 34950
    },
    {
      "epoch": 8.320180920073796,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 34951
    },
    {
      "epoch": 8.320418972802475,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 34952
    },
    {
      "epoch": 8.320657025531155,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0216,
      "step": 34953
    },
    {
      "epoch": 8.320895078259834,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 34954
    },
    {
      "epoch": 8.321133130988514,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 34955
    },
    {
      "epoch": 8.321371183717194,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 34956
    },
    {
      "epoch": 8.321609236445873,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0799,
      "step": 34957
    },
    {
      "epoch": 8.321847289174553,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2768,
      "step": 34958
    },
    {
      "epoch": 8.322085341903232,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 34959
    },
    {
      "epoch": 8.32232339463191,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 34960
    },
    {
      "epoch": 8.32256144736059,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1471,
      "step": 34961
    },
    {
      "epoch": 8.32279950008927,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0784,
      "step": 34962
    },
    {
      "epoch": 8.323037552817949,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0884,
      "step": 34963
    },
    {
      "epoch": 8.323275605546629,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 34964
    },
    {
      "epoch": 8.323513658275308,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 34965
    },
    {
      "epoch": 8.323751711003988,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 34966
    },
    {
      "epoch": 8.323989763732667,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 34967
    },
    {
      "epoch": 8.324227816461347,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 34968
    },
    {
      "epoch": 8.324465869190025,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9667,
      "step": 34969
    },
    {
      "epoch": 8.324703921918704,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0226,
      "step": 34970
    },
    {
      "epoch": 8.324941974647384,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2608,
      "step": 34971
    },
    {
      "epoch": 8.325180027376064,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7228,
      "step": 34972
    },
    {
      "epoch": 8.325418080104743,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 34973
    },
    {
      "epoch": 8.325656132833423,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 34974
    },
    {
      "epoch": 8.325894185562102,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 34975
    },
    {
      "epoch": 8.326132238290782,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 34976
    },
    {
      "epoch": 8.326370291019462,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 34977
    },
    {
      "epoch": 8.32660834374814,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1761,
      "step": 34978
    },
    {
      "epoch": 8.32684639647682,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 34979
    },
    {
      "epoch": 8.327084449205499,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2715,
      "step": 34980
    },
    {
      "epoch": 8.327322501934178,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.809,
      "step": 34981
    },
    {
      "epoch": 8.327560554662858,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 34982
    },
    {
      "epoch": 8.327798607391538,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 34983
    },
    {
      "epoch": 8.328036660120217,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 34984
    },
    {
      "epoch": 8.328274712848897,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 34985
    },
    {
      "epoch": 8.328512765577575,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 34986
    },
    {
      "epoch": 8.328750818306254,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7333,
      "step": 34987
    },
    {
      "epoch": 8.328988871034934,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 34988
    },
    {
      "epoch": 8.329226923763613,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 34989
    },
    {
      "epoch": 8.329464976492293,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9811,
      "step": 34990
    },
    {
      "epoch": 8.329703029220973,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 34991
    },
    {
      "epoch": 8.329941081949652,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8821,
      "step": 34992
    },
    {
      "epoch": 8.330179134678332,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 34993
    },
    {
      "epoch": 8.330417187407011,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 34994
    },
    {
      "epoch": 8.33065524013569,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 34995
    },
    {
      "epoch": 8.330893292864369,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 34996
    },
    {
      "epoch": 8.331131345593048,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 34997
    },
    {
      "epoch": 8.331369398321728,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9167,
      "step": 34998
    },
    {
      "epoch": 8.331607451050408,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2255,
      "step": 34999
    },
    {
      "epoch": 8.331845503779087,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 35000
    },
    {
      "epoch": 8.331845503779087,
      "eval_runtime": 0.511,
      "eval_samples_per_second": 1.957,
      "eval_steps_per_second": 1.957,
      "step": 35000
    },
    {
      "epoch": 8.332083556507767,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2045,
      "step": 35001
    },
    {
      "epoch": 8.332321609236446,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 35002
    },
    {
      "epoch": 8.332559661965126,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 35003
    },
    {
      "epoch": 8.332797714693804,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 35004
    },
    {
      "epoch": 8.333035767422484,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1778,
      "step": 35005
    },
    {
      "epoch": 8.333273820151163,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 35006
    },
    {
      "epoch": 8.333511872879843,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.82,
      "step": 35007
    },
    {
      "epoch": 8.333749925608522,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7938,
      "step": 35008
    },
    {
      "epoch": 8.333987978337202,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 35009
    },
    {
      "epoch": 8.334226031065882,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0835,
      "step": 35010
    },
    {
      "epoch": 8.334464083794561,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3668,
      "step": 35011
    },
    {
      "epoch": 8.33470213652324,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7161,
      "step": 35012
    },
    {
      "epoch": 8.334940189251919,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2367,
      "step": 35013
    },
    {
      "epoch": 8.335178241980598,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2826,
      "step": 35014
    },
    {
      "epoch": 8.335416294709278,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8484,
      "step": 35015
    },
    {
      "epoch": 8.335654347437957,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 35016
    },
    {
      "epoch": 8.335892400166637,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9523,
      "step": 35017
    },
    {
      "epoch": 8.336130452895317,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 35018
    },
    {
      "epoch": 8.336368505623996,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2044,
      "step": 35019
    },
    {
      "epoch": 8.336606558352676,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0705,
      "step": 35020
    },
    {
      "epoch": 8.336844611081354,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 35021
    },
    {
      "epoch": 8.337082663810033,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 35022
    },
    {
      "epoch": 8.337320716538713,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4339,
      "step": 35023
    },
    {
      "epoch": 8.337558769267392,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 35024
    },
    {
      "epoch": 8.337796821996072,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 35025
    },
    {
      "epoch": 8.338034874724752,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0071,
      "step": 35026
    },
    {
      "epoch": 8.338272927453431,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2741,
      "step": 35027
    },
    {
      "epoch": 8.338510980182111,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 35028
    },
    {
      "epoch": 8.33874903291079,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8533,
      "step": 35029
    },
    {
      "epoch": 8.338987085639468,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 35030
    },
    {
      "epoch": 8.339225138368148,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9243,
      "step": 35031
    },
    {
      "epoch": 8.339463191096828,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3802,
      "step": 35032
    },
    {
      "epoch": 8.339701243825507,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 35033
    },
    {
      "epoch": 8.339939296554187,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 35034
    },
    {
      "epoch": 8.340177349282866,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 35035
    },
    {
      "epoch": 8.340415402011546,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2001,
      "step": 35036
    },
    {
      "epoch": 8.340653454740226,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9724,
      "step": 35037
    },
    {
      "epoch": 8.340891507468905,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 35038
    },
    {
      "epoch": 8.341129560197583,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 35039
    },
    {
      "epoch": 8.341367612926263,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7993,
      "step": 35040
    },
    {
      "epoch": 8.341605665654942,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1889,
      "step": 35041
    },
    {
      "epoch": 8.341843718383622,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2375,
      "step": 35042
    },
    {
      "epoch": 8.342081771112301,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.6916,
      "step": 35043
    },
    {
      "epoch": 8.342319823840981,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 35044
    },
    {
      "epoch": 8.34255787656966,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9147,
      "step": 35045
    },
    {
      "epoch": 8.34279592929834,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 1.8191,
      "step": 35046
    },
    {
      "epoch": 8.34303398202702,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 35047
    },
    {
      "epoch": 8.343272034755698,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 35048
    },
    {
      "epoch": 8.343510087484377,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 35049
    },
    {
      "epoch": 8.343748140213057,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 35050
    },
    {
      "epoch": 8.343986192941736,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 35051
    },
    {
      "epoch": 8.344224245670416,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 35052
    },
    {
      "epoch": 8.344462298399096,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0169,
      "step": 35053
    },
    {
      "epoch": 8.344700351127775,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 35054
    },
    {
      "epoch": 8.344938403856455,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 35055
    },
    {
      "epoch": 8.345176456585133,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 35056
    },
    {
      "epoch": 8.345414509313812,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3881,
      "step": 35057
    },
    {
      "epoch": 8.345652562042492,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 35058
    },
    {
      "epoch": 8.345890614771172,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 35059
    },
    {
      "epoch": 8.346128667499851,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 35060
    },
    {
      "epoch": 8.34636672022853,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 35061
    },
    {
      "epoch": 8.34660477295721,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 35062
    },
    {
      "epoch": 8.34684282568589,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2706,
      "step": 35063
    },
    {
      "epoch": 8.34708087841457,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 35064
    },
    {
      "epoch": 8.347318931143247,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9889,
      "step": 35065
    },
    {
      "epoch": 8.347556983871927,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2203,
      "step": 35066
    },
    {
      "epoch": 8.347795036600607,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 35067
    },
    {
      "epoch": 8.348033089329286,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1031,
      "step": 35068
    },
    {
      "epoch": 8.348271142057966,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 35069
    },
    {
      "epoch": 8.348509194786645,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 35070
    },
    {
      "epoch": 8.348747247515325,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2857,
      "step": 35071
    },
    {
      "epoch": 8.348985300244005,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 35072
    },
    {
      "epoch": 8.349223352972684,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 35073
    },
    {
      "epoch": 8.349461405701362,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 35074
    },
    {
      "epoch": 8.349699458430042,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 35075
    },
    {
      "epoch": 8.349937511158721,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 35076
    },
    {
      "epoch": 8.350175563887401,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 35077
    },
    {
      "epoch": 8.35041361661608,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 35078
    },
    {
      "epoch": 8.35065166934476,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2298,
      "step": 35079
    },
    {
      "epoch": 8.35088972207344,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.5929,
      "step": 35080
    },
    {
      "epoch": 8.35112777480212,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 35081
    },
    {
      "epoch": 8.351365827530799,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0922,
      "step": 35082
    },
    {
      "epoch": 8.351603880259477,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9166,
      "step": 35083
    },
    {
      "epoch": 8.351841932988156,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 35084
    },
    {
      "epoch": 8.352079985716836,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 35085
    },
    {
      "epoch": 8.352318038445516,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 35086
    },
    {
      "epoch": 8.352556091174195,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2604,
      "step": 35087
    },
    {
      "epoch": 8.352794143902875,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 35088
    },
    {
      "epoch": 8.353032196631554,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 35089
    },
    {
      "epoch": 8.353270249360234,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 35090
    },
    {
      "epoch": 8.353508302088912,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9844,
      "step": 35091
    },
    {
      "epoch": 8.353746354817591,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 35092
    },
    {
      "epoch": 8.353984407546271,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 35093
    },
    {
      "epoch": 8.35422246027495,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 35094
    },
    {
      "epoch": 8.35446051300363,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 35095
    },
    {
      "epoch": 8.35469856573231,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4775,
      "step": 35096
    },
    {
      "epoch": 8.35493661846099,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8416,
      "step": 35097
    },
    {
      "epoch": 8.355174671189669,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2659,
      "step": 35098
    },
    {
      "epoch": 8.355412723918349,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2113,
      "step": 35099
    },
    {
      "epoch": 8.355650776647026,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 35100
    },
    {
      "epoch": 8.355650776647026,
      "eval_runtime": 0.5218,
      "eval_samples_per_second": 1.916,
      "eval_steps_per_second": 1.916,
      "step": 35100
    },
    {
      "epoch": 8.355888829375706,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 35101
    },
    {
      "epoch": 8.356126882104386,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 35102
    },
    {
      "epoch": 8.356364934833065,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 35103
    },
    {
      "epoch": 8.356602987561745,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9072,
      "step": 35104
    },
    {
      "epoch": 8.356841040290425,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 35105
    },
    {
      "epoch": 8.357079093019104,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 35106
    },
    {
      "epoch": 8.357317145747784,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1502,
      "step": 35107
    },
    {
      "epoch": 8.357555198476463,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1756,
      "step": 35108
    },
    {
      "epoch": 8.357793251205141,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 35109
    },
    {
      "epoch": 8.35803130393382,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 35110
    },
    {
      "epoch": 8.3582693566625,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0842,
      "step": 35111
    },
    {
      "epoch": 8.35850740939118,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 35112
    },
    {
      "epoch": 8.35874546211986,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1419,
      "step": 35113
    },
    {
      "epoch": 8.35898351484854,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 35114
    },
    {
      "epoch": 8.359221567577219,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 35115
    },
    {
      "epoch": 8.359459620305898,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 35116
    },
    {
      "epoch": 8.359697673034578,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 35117
    },
    {
      "epoch": 8.359935725763256,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0065,
      "step": 35118
    },
    {
      "epoch": 8.360173778491935,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2887,
      "step": 35119
    },
    {
      "epoch": 8.360411831220615,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6239,
      "step": 35120
    },
    {
      "epoch": 8.360649883949295,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 35121
    },
    {
      "epoch": 8.360887936677974,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 35122
    },
    {
      "epoch": 8.361125989406654,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0669,
      "step": 35123
    },
    {
      "epoch": 8.361364042135333,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 35124
    },
    {
      "epoch": 8.361602094864013,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0002,
      "step": 35125
    },
    {
      "epoch": 8.361840147592691,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1857,
      "step": 35126
    },
    {
      "epoch": 8.36207820032137,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3385,
      "step": 35127
    },
    {
      "epoch": 8.36231625305005,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0793,
      "step": 35128
    },
    {
      "epoch": 8.36255430577873,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 35129
    },
    {
      "epoch": 8.36279235850741,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 35130
    },
    {
      "epoch": 8.363030411236089,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 35131
    },
    {
      "epoch": 8.363268463964769,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 35132
    },
    {
      "epoch": 8.363506516693448,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 35133
    },
    {
      "epoch": 8.363744569422128,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 35134
    },
    {
      "epoch": 8.363982622150806,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 35135
    },
    {
      "epoch": 8.364220674879485,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8941,
      "step": 35136
    },
    {
      "epoch": 8.364458727608165,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7708,
      "step": 35137
    },
    {
      "epoch": 8.364696780336844,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3343,
      "step": 35138
    },
    {
      "epoch": 8.364934833065524,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1469,
      "step": 35139
    },
    {
      "epoch": 8.365172885794204,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 35140
    },
    {
      "epoch": 8.365410938522883,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 35141
    },
    {
      "epoch": 8.365648991251563,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4095,
      "step": 35142
    },
    {
      "epoch": 8.365887043980242,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1648,
      "step": 35143
    },
    {
      "epoch": 8.36612509670892,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 35144
    },
    {
      "epoch": 8.3663631494376,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 35145
    },
    {
      "epoch": 8.36660120216628,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 35146
    },
    {
      "epoch": 8.366839254894959,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 35147
    },
    {
      "epoch": 8.367077307623639,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 35148
    },
    {
      "epoch": 8.367315360352318,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 35149
    },
    {
      "epoch": 8.367553413080998,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0976,
      "step": 35150
    },
    {
      "epoch": 8.367791465809677,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3339,
      "step": 35151
    },
    {
      "epoch": 8.368029518538357,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9583,
      "step": 35152
    },
    {
      "epoch": 8.368267571267035,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 2.0449,
      "step": 35153
    },
    {
      "epoch": 8.368505623995715,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 35154
    },
    {
      "epoch": 8.368743676724394,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 35155
    },
    {
      "epoch": 8.368981729453074,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3107,
      "step": 35156
    },
    {
      "epoch": 8.369219782181753,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 35157
    },
    {
      "epoch": 8.369457834910433,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 35158
    },
    {
      "epoch": 8.369695887639113,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 35159
    },
    {
      "epoch": 8.369933940367792,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 35160
    },
    {
      "epoch": 8.37017199309647,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7538,
      "step": 35161
    },
    {
      "epoch": 8.37041004582515,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 35162
    },
    {
      "epoch": 8.37064809855383,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 35163
    },
    {
      "epoch": 8.370886151282509,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 35164
    },
    {
      "epoch": 8.371124204011188,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 35165
    },
    {
      "epoch": 8.371362256739868,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3626,
      "step": 35166
    },
    {
      "epoch": 8.371600309468548,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 35167
    },
    {
      "epoch": 8.371838362197227,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 35168
    },
    {
      "epoch": 8.372076414925907,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 35169
    },
    {
      "epoch": 8.372314467654585,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0355,
      "step": 35170
    },
    {
      "epoch": 8.372552520383264,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 35171
    },
    {
      "epoch": 8.372790573111944,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 35172
    },
    {
      "epoch": 8.373028625840623,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2503,
      "step": 35173
    },
    {
      "epoch": 8.373266678569303,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 35174
    },
    {
      "epoch": 8.373504731297983,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 35175
    },
    {
      "epoch": 8.373742784026662,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0084,
      "step": 35176
    },
    {
      "epoch": 8.373980836755342,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7482,
      "step": 35177
    },
    {
      "epoch": 8.374218889484021,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.232,
      "step": 35178
    },
    {
      "epoch": 8.3744569422127,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 35179
    },
    {
      "epoch": 8.374694994941379,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 35180
    },
    {
      "epoch": 8.374933047670059,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 35181
    },
    {
      "epoch": 8.375171100398738,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 35182
    },
    {
      "epoch": 8.375409153127418,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3816,
      "step": 35183
    },
    {
      "epoch": 8.375647205856097,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1719,
      "step": 35184
    },
    {
      "epoch": 8.375885258584777,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1791,
      "step": 35185
    },
    {
      "epoch": 8.376123311313457,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 35186
    },
    {
      "epoch": 8.376361364042136,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 35187
    },
    {
      "epoch": 8.376599416770814,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8017,
      "step": 35188
    },
    {
      "epoch": 8.376837469499494,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7255,
      "step": 35189
    },
    {
      "epoch": 8.377075522228173,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 35190
    },
    {
      "epoch": 8.377313574956853,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 35191
    },
    {
      "epoch": 8.377551627685532,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.6155,
      "step": 35192
    },
    {
      "epoch": 8.377789680414212,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2691,
      "step": 35193
    },
    {
      "epoch": 8.378027733142892,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 35194
    },
    {
      "epoch": 8.378265785871571,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 35195
    },
    {
      "epoch": 8.378503838600249,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1356,
      "step": 35196
    },
    {
      "epoch": 8.378741891328929,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 35197
    },
    {
      "epoch": 8.378979944057608,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 35198
    },
    {
      "epoch": 8.379217996786288,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 35199
    },
    {
      "epoch": 8.379456049514967,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 35200
    },
    {
      "epoch": 8.379456049514967,
      "eval_runtime": 0.5188,
      "eval_samples_per_second": 1.927,
      "eval_steps_per_second": 1.927,
      "step": 35200
    },
    {
      "epoch": 8.379694102243647,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 35201
    },
    {
      "epoch": 8.379932154972327,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 35202
    },
    {
      "epoch": 8.380170207701006,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 35203
    },
    {
      "epoch": 8.380408260429686,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8618,
      "step": 35204
    },
    {
      "epoch": 8.380646313158364,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 35205
    },
    {
      "epoch": 8.380884365887043,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 35206
    },
    {
      "epoch": 8.381122418615723,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7269,
      "step": 35207
    },
    {
      "epoch": 8.381360471344403,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 35208
    },
    {
      "epoch": 8.381598524073082,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.223,
      "step": 35209
    },
    {
      "epoch": 8.381836576801762,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 35210
    },
    {
      "epoch": 8.382074629530441,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 35211
    },
    {
      "epoch": 8.382312682259121,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9159,
      "step": 35212
    },
    {
      "epoch": 8.3825507349878,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1605,
      "step": 35213
    },
    {
      "epoch": 8.382788787716478,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 35214
    },
    {
      "epoch": 8.383026840445158,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 35215
    },
    {
      "epoch": 8.383264893173838,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 35216
    },
    {
      "epoch": 8.383502945902517,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 35217
    },
    {
      "epoch": 8.383740998631197,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2272,
      "step": 35218
    },
    {
      "epoch": 8.383979051359876,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 35219
    },
    {
      "epoch": 8.384217104088556,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4569,
      "step": 35220
    },
    {
      "epoch": 8.384455156817236,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 35221
    },
    {
      "epoch": 8.384693209545915,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 35222
    },
    {
      "epoch": 8.384931262274593,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 35223
    },
    {
      "epoch": 8.385169315003273,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1581,
      "step": 35224
    },
    {
      "epoch": 8.385407367731952,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 35225
    },
    {
      "epoch": 8.385645420460632,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2582,
      "step": 35226
    },
    {
      "epoch": 8.385883473189311,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 35227
    },
    {
      "epoch": 8.386121525917991,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 35228
    },
    {
      "epoch": 8.38635957864667,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 35229
    },
    {
      "epoch": 8.38659763137535,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 35230
    },
    {
      "epoch": 8.386835684104028,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 35231
    },
    {
      "epoch": 8.387073736832708,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 35232
    },
    {
      "epoch": 8.387311789561387,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.888,
      "step": 35233
    },
    {
      "epoch": 8.387549842290067,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 35234
    },
    {
      "epoch": 8.387787895018747,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 35235
    },
    {
      "epoch": 8.388025947747426,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3628,
      "step": 35236
    },
    {
      "epoch": 8.388264000476106,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 35237
    },
    {
      "epoch": 8.388502053204785,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 35238
    },
    {
      "epoch": 8.388740105933465,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 35239
    },
    {
      "epoch": 8.388978158662145,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 35240
    },
    {
      "epoch": 8.389216211390822,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 35241
    },
    {
      "epoch": 8.389454264119502,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 35242
    },
    {
      "epoch": 8.389692316848182,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 35243
    },
    {
      "epoch": 8.389930369576861,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 35244
    },
    {
      "epoch": 8.39016842230554,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.482,
      "step": 35245
    },
    {
      "epoch": 8.39040647503422,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 35246
    },
    {
      "epoch": 8.3906445277629,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 35247
    },
    {
      "epoch": 8.39088258049158,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8414,
      "step": 35248
    },
    {
      "epoch": 8.391120633220257,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 35249
    },
    {
      "epoch": 8.391358685948937,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.913,
      "step": 35250
    },
    {
      "epoch": 8.391596738677617,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.6598,
      "step": 35251
    },
    {
      "epoch": 8.391834791406296,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 35252
    },
    {
      "epoch": 8.392072844134976,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 35253
    },
    {
      "epoch": 8.392310896863656,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2781,
      "step": 35254
    },
    {
      "epoch": 8.392548949592335,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 35255
    },
    {
      "epoch": 8.392787002321015,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 35256
    },
    {
      "epoch": 8.393025055049694,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 35257
    },
    {
      "epoch": 8.393263107778372,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1793,
      "step": 35258
    },
    {
      "epoch": 8.393501160507052,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 35259
    },
    {
      "epoch": 8.393739213235731,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 35260
    },
    {
      "epoch": 8.393977265964411,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2177,
      "step": 35261
    },
    {
      "epoch": 8.39421531869309,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 35262
    },
    {
      "epoch": 8.39445337142177,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 35263
    },
    {
      "epoch": 8.39469142415045,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4623,
      "step": 35264
    },
    {
      "epoch": 8.39492947687913,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 35265
    },
    {
      "epoch": 8.395167529607809,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 35266
    },
    {
      "epoch": 8.395405582336487,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.5251,
      "step": 35267
    },
    {
      "epoch": 8.395643635065166,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 35268
    },
    {
      "epoch": 8.395881687793846,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.4257,
      "step": 35269
    },
    {
      "epoch": 8.396119740522526,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 35270
    },
    {
      "epoch": 8.396357793251205,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 35271
    },
    {
      "epoch": 8.396595845979885,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 1.9378,
      "step": 35272
    },
    {
      "epoch": 8.396833898708564,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 35273
    },
    {
      "epoch": 8.397071951437244,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 35274
    },
    {
      "epoch": 8.397310004165924,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 35275
    },
    {
      "epoch": 8.397548056894601,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.843,
      "step": 35276
    },
    {
      "epoch": 8.397786109623281,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 35277
    },
    {
      "epoch": 8.39802416235196,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1819,
      "step": 35278
    },
    {
      "epoch": 8.39826221508064,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 35279
    },
    {
      "epoch": 8.39850026780932,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.093,
      "step": 35280
    },
    {
      "epoch": 8.398738320538,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.493,
      "step": 35281
    },
    {
      "epoch": 8.39897637326668,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 35282
    },
    {
      "epoch": 8.399214425995359,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 35283
    },
    {
      "epoch": 8.399452478724037,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 35284
    },
    {
      "epoch": 8.399690531452716,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3023,
      "step": 35285
    },
    {
      "epoch": 8.399928584181396,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 35286
    },
    {
      "epoch": 8.400166636910075,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 35287
    },
    {
      "epoch": 8.400404689638755,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 35288
    },
    {
      "epoch": 8.400642742367435,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8763,
      "step": 35289
    },
    {
      "epoch": 8.400880795096114,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3904,
      "step": 35290
    },
    {
      "epoch": 8.401118847824794,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 35291
    },
    {
      "epoch": 8.401356900553473,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 35292
    },
    {
      "epoch": 8.401594953282151,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8809,
      "step": 35293
    },
    {
      "epoch": 8.40183300601083,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6618,
      "step": 35294
    },
    {
      "epoch": 8.40207105873951,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9235,
      "step": 35295
    },
    {
      "epoch": 8.40230911146819,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3274,
      "step": 35296
    },
    {
      "epoch": 8.40254716419687,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 35297
    },
    {
      "epoch": 8.40278521692555,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1678,
      "step": 35298
    },
    {
      "epoch": 8.403023269654229,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2994,
      "step": 35299
    },
    {
      "epoch": 8.403261322382908,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 35300
    },
    {
      "epoch": 8.403261322382908,
      "eval_runtime": 0.521,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 35300
    },
    {
      "epoch": 8.403499375111588,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 35301
    },
    {
      "epoch": 8.403737427840266,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3229,
      "step": 35302
    },
    {
      "epoch": 8.403975480568945,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0404,
      "step": 35303
    },
    {
      "epoch": 8.404213533297625,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.6772,
      "step": 35304
    },
    {
      "epoch": 8.404451586026305,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 35305
    },
    {
      "epoch": 8.404689638754984,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 35306
    },
    {
      "epoch": 8.404927691483664,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0191,
      "step": 35307
    },
    {
      "epoch": 8.405165744212344,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7502,
      "step": 35308
    },
    {
      "epoch": 8.405403796941023,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7679,
      "step": 35309
    },
    {
      "epoch": 8.405641849669703,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 35310
    },
    {
      "epoch": 8.40587990239838,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2227,
      "step": 35311
    },
    {
      "epoch": 8.40611795512706,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0702,
      "step": 35312
    },
    {
      "epoch": 8.40635600785574,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2011,
      "step": 35313
    },
    {
      "epoch": 8.40659406058442,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3075,
      "step": 35314
    },
    {
      "epoch": 8.406832113313099,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2735,
      "step": 35315
    },
    {
      "epoch": 8.407070166041779,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 35316
    },
    {
      "epoch": 8.407308218770458,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8073,
      "step": 35317
    },
    {
      "epoch": 8.407546271499138,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 35318
    },
    {
      "epoch": 8.407784324227816,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 35319
    },
    {
      "epoch": 8.408022376956495,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3501,
      "step": 35320
    },
    {
      "epoch": 8.408260429685175,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 35321
    },
    {
      "epoch": 8.408498482413854,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7683,
      "step": 35322
    },
    {
      "epoch": 8.408736535142534,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3559,
      "step": 35323
    },
    {
      "epoch": 8.408974587871214,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 35324
    },
    {
      "epoch": 8.409212640599893,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 35325
    },
    {
      "epoch": 8.409450693328573,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 35326
    },
    {
      "epoch": 8.409688746057252,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 35327
    },
    {
      "epoch": 8.40992679878593,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1878,
      "step": 35328
    },
    {
      "epoch": 8.41016485151461,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 35329
    },
    {
      "epoch": 8.41040290424329,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 35330
    },
    {
      "epoch": 8.41064095697197,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1981,
      "step": 35331
    },
    {
      "epoch": 8.410879009700649,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1012,
      "step": 35332
    },
    {
      "epoch": 8.411117062429328,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 35333
    },
    {
      "epoch": 8.411355115158008,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 35334
    },
    {
      "epoch": 8.411593167886688,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 35335
    },
    {
      "epoch": 8.411831220615367,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 35336
    },
    {
      "epoch": 8.412069273344045,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2722,
      "step": 35337
    },
    {
      "epoch": 8.412307326072725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 35338
    },
    {
      "epoch": 8.412545378801404,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 35339
    },
    {
      "epoch": 8.412783431530084,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.8003,
      "step": 35340
    },
    {
      "epoch": 8.413021484258763,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 35341
    },
    {
      "epoch": 8.413259536987443,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2983,
      "step": 35342
    },
    {
      "epoch": 8.413497589716123,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0375,
      "step": 35343
    },
    {
      "epoch": 8.413735642444802,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.012,
      "step": 35344
    },
    {
      "epoch": 8.413973695173482,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0786,
      "step": 35345
    },
    {
      "epoch": 8.41421174790216,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 35346
    },
    {
      "epoch": 8.41444980063084,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7508,
      "step": 35347
    },
    {
      "epoch": 8.414687853359519,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3185,
      "step": 35348
    },
    {
      "epoch": 8.414925906088198,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0928,
      "step": 35349
    },
    {
      "epoch": 8.415163958816878,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 35350
    },
    {
      "epoch": 8.415402011545558,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 35351
    },
    {
      "epoch": 8.415640064274237,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 35352
    },
    {
      "epoch": 8.415878117002917,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 35353
    },
    {
      "epoch": 8.416116169731595,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0556,
      "step": 35354
    },
    {
      "epoch": 8.416354222460274,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1073,
      "step": 35355
    },
    {
      "epoch": 8.416592275188954,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8644,
      "step": 35356
    },
    {
      "epoch": 8.416830327917634,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 35357
    },
    {
      "epoch": 8.417068380646313,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 35358
    },
    {
      "epoch": 8.417306433374993,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 35359
    },
    {
      "epoch": 8.417544486103672,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 35360
    },
    {
      "epoch": 8.417782538832352,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9023,
      "step": 35361
    },
    {
      "epoch": 8.418020591561032,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 35362
    },
    {
      "epoch": 8.41825864428971,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 35363
    },
    {
      "epoch": 8.418496697018389,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 35364
    },
    {
      "epoch": 8.418734749747069,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 35365
    },
    {
      "epoch": 8.418972802475748,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 35366
    },
    {
      "epoch": 8.419210855204428,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 35367
    },
    {
      "epoch": 8.419448907933107,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 35368
    },
    {
      "epoch": 8.419686960661787,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 35369
    },
    {
      "epoch": 8.419925013390467,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 35370
    },
    {
      "epoch": 8.420163066119146,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.3964,
      "step": 35371
    },
    {
      "epoch": 8.420401118847824,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 35372
    },
    {
      "epoch": 8.420639171576504,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 35373
    },
    {
      "epoch": 8.420877224305183,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 35374
    },
    {
      "epoch": 8.421115277033863,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 35375
    },
    {
      "epoch": 8.421353329762542,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 35376
    },
    {
      "epoch": 8.421591382491222,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 35377
    },
    {
      "epoch": 8.421829435219902,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 35378
    },
    {
      "epoch": 8.422067487948581,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 35379
    },
    {
      "epoch": 8.422305540677261,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8428,
      "step": 35380
    },
    {
      "epoch": 8.422543593405939,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 35381
    },
    {
      "epoch": 8.422781646134618,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9526,
      "step": 35382
    },
    {
      "epoch": 8.423019698863298,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 35383
    },
    {
      "epoch": 8.423257751591978,
      "grad_norm": 6.65625,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 35384
    },
    {
      "epoch": 8.423495804320657,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 35385
    },
    {
      "epoch": 8.423733857049337,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1451,
      "step": 35386
    },
    {
      "epoch": 8.423971909778016,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 35387
    },
    {
      "epoch": 8.424209962506696,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 35388
    },
    {
      "epoch": 8.424448015235374,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 35389
    },
    {
      "epoch": 8.424686067964053,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 35390
    },
    {
      "epoch": 8.424924120692733,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 35391
    },
    {
      "epoch": 8.425162173421413,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 35392
    },
    {
      "epoch": 8.425400226150092,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 35393
    },
    {
      "epoch": 8.425638278878772,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 35394
    },
    {
      "epoch": 8.425876331607451,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 35395
    },
    {
      "epoch": 8.426114384336131,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.7133,
      "step": 35396
    },
    {
      "epoch": 8.42635243706481,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9978,
      "step": 35397
    },
    {
      "epoch": 8.426590489793488,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6632,
      "step": 35398
    },
    {
      "epoch": 8.426828542522168,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 35399
    },
    {
      "epoch": 8.427066595250848,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8462,
      "step": 35400
    },
    {
      "epoch": 8.427066595250848,
      "eval_runtime": 0.5238,
      "eval_samples_per_second": 1.909,
      "eval_steps_per_second": 1.909,
      "step": 35400
    },
    {
      "epoch": 8.427304647979527,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8697,
      "step": 35401
    },
    {
      "epoch": 8.427542700708207,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3392,
      "step": 35402
    },
    {
      "epoch": 8.427780753436886,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 35403
    },
    {
      "epoch": 8.428018806165566,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1104,
      "step": 35404
    },
    {
      "epoch": 8.428256858894246,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2301,
      "step": 35405
    },
    {
      "epoch": 8.428494911622925,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 35406
    },
    {
      "epoch": 8.428732964351603,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.244,
      "step": 35407
    },
    {
      "epoch": 8.428971017080283,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 35408
    },
    {
      "epoch": 8.429209069808962,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8979,
      "step": 35409
    },
    {
      "epoch": 8.429447122537642,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 35410
    },
    {
      "epoch": 8.429685175266322,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2354,
      "step": 35411
    },
    {
      "epoch": 8.429923227995001,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 35412
    },
    {
      "epoch": 8.43016128072368,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1817,
      "step": 35413
    },
    {
      "epoch": 8.43039933345236,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1929,
      "step": 35414
    },
    {
      "epoch": 8.43063738618104,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 35415
    },
    {
      "epoch": 8.430875438909718,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7909,
      "step": 35416
    },
    {
      "epoch": 8.431113491638397,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 35417
    },
    {
      "epoch": 8.431351544367077,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 35418
    },
    {
      "epoch": 8.431589597095757,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 35419
    },
    {
      "epoch": 8.431827649824436,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 35420
    },
    {
      "epoch": 8.432065702553116,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8149,
      "step": 35421
    },
    {
      "epoch": 8.432303755281795,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 35422
    },
    {
      "epoch": 8.432541808010475,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0409,
      "step": 35423
    },
    {
      "epoch": 8.432779860739153,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.189,
      "step": 35424
    },
    {
      "epoch": 8.433017913467832,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 35425
    },
    {
      "epoch": 8.433255966196512,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 35426
    },
    {
      "epoch": 8.433494018925192,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 35427
    },
    {
      "epoch": 8.433732071653871,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 35428
    },
    {
      "epoch": 8.433970124382551,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 35429
    },
    {
      "epoch": 8.43420817711123,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 35430
    },
    {
      "epoch": 8.43444622983991,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9606,
      "step": 35431
    },
    {
      "epoch": 8.43468428256859,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 35432
    },
    {
      "epoch": 8.434922335297268,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1669,
      "step": 35433
    },
    {
      "epoch": 8.435160388025947,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 35434
    },
    {
      "epoch": 8.435398440754627,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 35435
    },
    {
      "epoch": 8.435636493483306,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 35436
    },
    {
      "epoch": 8.435874546211986,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 35437
    },
    {
      "epoch": 8.436112598940666,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 35438
    },
    {
      "epoch": 8.436350651669345,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 35439
    },
    {
      "epoch": 8.436588704398025,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.3145,
      "step": 35440
    },
    {
      "epoch": 8.436826757126704,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 35441
    },
    {
      "epoch": 8.437064809855382,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 35442
    },
    {
      "epoch": 8.437302862584062,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 35443
    },
    {
      "epoch": 8.437540915312741,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.7735,
      "step": 35444
    },
    {
      "epoch": 8.437778968041421,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8657,
      "step": 35445
    },
    {
      "epoch": 8.4380170207701,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 35446
    },
    {
      "epoch": 8.43825507349878,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 35447
    },
    {
      "epoch": 8.43849312622746,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 35448
    },
    {
      "epoch": 8.43873117895614,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9661,
      "step": 35449
    },
    {
      "epoch": 8.438969231684819,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 35450
    },
    {
      "epoch": 8.439207284413497,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 35451
    },
    {
      "epoch": 8.439445337142176,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 35452
    },
    {
      "epoch": 8.439683389870856,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 35453
    },
    {
      "epoch": 8.439921442599536,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 35454
    },
    {
      "epoch": 8.440159495328215,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 35455
    },
    {
      "epoch": 8.440397548056895,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 35456
    },
    {
      "epoch": 8.440635600785575,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 35457
    },
    {
      "epoch": 8.440873653514254,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 35458
    },
    {
      "epoch": 8.441111706242932,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1231,
      "step": 35459
    },
    {
      "epoch": 8.441349758971612,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 35460
    },
    {
      "epoch": 8.441587811700291,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 35461
    },
    {
      "epoch": 8.44182586442897,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 35462
    },
    {
      "epoch": 8.44206391715765,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 35463
    },
    {
      "epoch": 8.44230196988633,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 35464
    },
    {
      "epoch": 8.44254002261501,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1658,
      "step": 35465
    },
    {
      "epoch": 8.44277807534369,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 35466
    },
    {
      "epoch": 8.443016128072369,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 35467
    },
    {
      "epoch": 8.443254180801047,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0121,
      "step": 35468
    },
    {
      "epoch": 8.443492233529726,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0516,
      "step": 35469
    },
    {
      "epoch": 8.443730286258406,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 35470
    },
    {
      "epoch": 8.443968338987085,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 35471
    },
    {
      "epoch": 8.444206391715765,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0451,
      "step": 35472
    },
    {
      "epoch": 8.444444444444445,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 35473
    },
    {
      "epoch": 8.444682497173124,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2424,
      "step": 35474
    },
    {
      "epoch": 8.444920549901804,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 35475
    },
    {
      "epoch": 8.445158602630483,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 35476
    },
    {
      "epoch": 8.445396655359161,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 35477
    },
    {
      "epoch": 8.445634708087841,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 35478
    },
    {
      "epoch": 8.44587276081652,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 35479
    },
    {
      "epoch": 8.4461108135452,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1013,
      "step": 35480
    },
    {
      "epoch": 8.44634886627388,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 35481
    },
    {
      "epoch": 8.44658691900256,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 35482
    },
    {
      "epoch": 8.446824971731239,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 35483
    },
    {
      "epoch": 8.447063024459919,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.242,
      "step": 35484
    },
    {
      "epoch": 8.447301077188598,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.1872,
      "step": 35485
    },
    {
      "epoch": 8.447539129917276,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 35486
    },
    {
      "epoch": 8.447777182645956,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 35487
    },
    {
      "epoch": 8.448015235374635,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2654,
      "step": 35488
    },
    {
      "epoch": 8.448253288103315,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 35489
    },
    {
      "epoch": 8.448491340831994,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 35490
    },
    {
      "epoch": 8.448729393560674,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 35491
    },
    {
      "epoch": 8.448967446289354,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.3159,
      "step": 35492
    },
    {
      "epoch": 8.449205499018033,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1644,
      "step": 35493
    },
    {
      "epoch": 8.449443551746711,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3252,
      "step": 35494
    },
    {
      "epoch": 8.44968160447539,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 35495
    },
    {
      "epoch": 8.44991965720407,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 35496
    },
    {
      "epoch": 8.45015770993275,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.5231,
      "step": 35497
    },
    {
      "epoch": 8.45039576266143,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 35498
    },
    {
      "epoch": 8.450633815390109,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9407,
      "step": 35499
    },
    {
      "epoch": 8.450871868118789,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 35500
    },
    {
      "epoch": 8.450871868118789,
      "eval_runtime": 0.5199,
      "eval_samples_per_second": 1.924,
      "eval_steps_per_second": 1.924,
      "step": 35500
    },
    {
      "epoch": 8.451109920847468,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 35501
    },
    {
      "epoch": 8.451347973576148,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3019,
      "step": 35502
    },
    {
      "epoch": 8.451586026304826,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 35503
    },
    {
      "epoch": 8.451824079033505,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4217,
      "step": 35504
    },
    {
      "epoch": 8.452062131762185,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1344,
      "step": 35505
    },
    {
      "epoch": 8.452300184490865,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 35506
    },
    {
      "epoch": 8.452538237219544,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3162,
      "step": 35507
    },
    {
      "epoch": 8.452776289948224,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 35508
    },
    {
      "epoch": 8.453014342676903,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 35509
    },
    {
      "epoch": 8.453252395405583,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.5124,
      "step": 35510
    },
    {
      "epoch": 8.453490448134263,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.6704,
      "step": 35511
    },
    {
      "epoch": 8.45372850086294,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9907,
      "step": 35512
    },
    {
      "epoch": 8.45396655359162,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 35513
    },
    {
      "epoch": 8.4542046063203,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8897,
      "step": 35514
    },
    {
      "epoch": 8.45444265904898,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8869,
      "step": 35515
    },
    {
      "epoch": 8.454680711777659,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2495,
      "step": 35516
    },
    {
      "epoch": 8.454918764506338,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4031,
      "step": 35517
    },
    {
      "epoch": 8.455156817235018,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 35518
    },
    {
      "epoch": 8.455394869963698,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 35519
    },
    {
      "epoch": 8.455632922692377,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 35520
    },
    {
      "epoch": 8.455870975421055,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 35521
    },
    {
      "epoch": 8.456109028149735,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8107,
      "step": 35522
    },
    {
      "epoch": 8.456347080878414,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 35523
    },
    {
      "epoch": 8.456585133607094,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 35524
    },
    {
      "epoch": 8.456823186335773,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 35525
    },
    {
      "epoch": 8.457061239064453,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 35526
    },
    {
      "epoch": 8.457299291793133,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 35527
    },
    {
      "epoch": 8.457537344521812,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 35528
    },
    {
      "epoch": 8.45777539725049,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4317,
      "step": 35529
    },
    {
      "epoch": 8.45801344997917,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 35530
    },
    {
      "epoch": 8.45825150270785,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.384,
      "step": 35531
    },
    {
      "epoch": 8.458489555436529,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 35532
    },
    {
      "epoch": 8.458727608165209,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 35533
    },
    {
      "epoch": 8.458965660893888,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 35534
    },
    {
      "epoch": 8.459203713622568,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 35535
    },
    {
      "epoch": 8.459441766351247,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 35536
    },
    {
      "epoch": 8.459679819079927,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 35537
    },
    {
      "epoch": 8.459917871808605,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 35538
    },
    {
      "epoch": 8.460155924537284,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8856,
      "step": 35539
    },
    {
      "epoch": 8.460393977265964,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9129,
      "step": 35540
    },
    {
      "epoch": 8.460632029994644,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9108,
      "step": 35541
    },
    {
      "epoch": 8.460870082723323,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 35542
    },
    {
      "epoch": 8.461108135452003,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 35543
    },
    {
      "epoch": 8.461346188180682,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 35544
    },
    {
      "epoch": 8.461584240909362,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0253,
      "step": 35545
    },
    {
      "epoch": 8.461822293638042,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 35546
    },
    {
      "epoch": 8.46206034636672,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 35547
    },
    {
      "epoch": 8.462298399095399,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 35548
    },
    {
      "epoch": 8.462536451824079,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 35549
    },
    {
      "epoch": 8.462774504552758,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 35550
    },
    {
      "epoch": 8.463012557281438,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 35551
    },
    {
      "epoch": 8.463250610010117,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0783,
      "step": 35552
    },
    {
      "epoch": 8.463488662738797,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 35553
    },
    {
      "epoch": 8.463726715467477,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9329,
      "step": 35554
    },
    {
      "epoch": 8.463964768196156,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.845,
      "step": 35555
    },
    {
      "epoch": 8.464202820924834,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 35556
    },
    {
      "epoch": 8.464440873653514,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 35557
    },
    {
      "epoch": 8.464678926382193,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 35558
    },
    {
      "epoch": 8.464916979110873,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3693,
      "step": 35559
    },
    {
      "epoch": 8.465155031839553,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3973,
      "step": 35560
    },
    {
      "epoch": 8.465393084568232,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0245,
      "step": 35561
    },
    {
      "epoch": 8.465631137296912,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 35562
    },
    {
      "epoch": 8.465869190025591,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 35563
    },
    {
      "epoch": 8.46610724275427,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 35564
    },
    {
      "epoch": 8.466345295482949,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 35565
    },
    {
      "epoch": 8.466583348211628,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.2303,
      "step": 35566
    },
    {
      "epoch": 8.466821400940308,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 35567
    },
    {
      "epoch": 8.467059453668988,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 35568
    },
    {
      "epoch": 8.467297506397667,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 35569
    },
    {
      "epoch": 8.467535559126347,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.483,
      "step": 35570
    },
    {
      "epoch": 8.467773611855026,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9366,
      "step": 35571
    },
    {
      "epoch": 8.468011664583706,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 35572
    },
    {
      "epoch": 8.468249717312384,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 35573
    },
    {
      "epoch": 8.468487770041063,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8415,
      "step": 35574
    },
    {
      "epoch": 8.468725822769743,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3308,
      "step": 35575
    },
    {
      "epoch": 8.468963875498423,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 35576
    },
    {
      "epoch": 8.469201928227102,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 35577
    },
    {
      "epoch": 8.469439980955782,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9628,
      "step": 35578
    },
    {
      "epoch": 8.469678033684461,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 35579
    },
    {
      "epoch": 8.469916086413141,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2747,
      "step": 35580
    },
    {
      "epoch": 8.47015413914182,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 35581
    },
    {
      "epoch": 8.470392191870499,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 35582
    },
    {
      "epoch": 8.470630244599178,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8382,
      "step": 35583
    },
    {
      "epoch": 8.470868297327858,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1409,
      "step": 35584
    },
    {
      "epoch": 8.471106350056537,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9541,
      "step": 35585
    },
    {
      "epoch": 8.471344402785217,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 35586
    },
    {
      "epoch": 8.471582455513897,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 35587
    },
    {
      "epoch": 8.471820508242576,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 35588
    },
    {
      "epoch": 8.472058560971256,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0474,
      "step": 35589
    },
    {
      "epoch": 8.472296613699935,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9182,
      "step": 35590
    },
    {
      "epoch": 8.472534666428613,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 35591
    },
    {
      "epoch": 8.472772719157293,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 35592
    },
    {
      "epoch": 8.473010771885972,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 35593
    },
    {
      "epoch": 8.473248824614652,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 35594
    },
    {
      "epoch": 8.473486877343332,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1306,
      "step": 35595
    },
    {
      "epoch": 8.473724930072011,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 35596
    },
    {
      "epoch": 8.47396298280069,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.7453,
      "step": 35597
    },
    {
      "epoch": 8.47420103552937,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 35598
    },
    {
      "epoch": 8.474439088258048,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 35599
    },
    {
      "epoch": 8.474677140986728,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 35600
    },
    {
      "epoch": 8.474677140986728,
      "eval_runtime": 0.5288,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 35600
    },
    {
      "epoch": 8.474915193715407,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 35601
    },
    {
      "epoch": 8.475153246444087,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 35602
    },
    {
      "epoch": 8.475391299172767,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 35603
    },
    {
      "epoch": 8.475629351901446,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 35604
    },
    {
      "epoch": 8.475867404630126,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1905,
      "step": 35605
    },
    {
      "epoch": 8.476105457358805,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 35606
    },
    {
      "epoch": 8.476343510087485,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 35607
    },
    {
      "epoch": 8.476581562816163,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 35608
    },
    {
      "epoch": 8.476819615544843,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 35609
    },
    {
      "epoch": 8.477057668273522,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3419,
      "step": 35610
    },
    {
      "epoch": 8.477295721002202,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3472,
      "step": 35611
    },
    {
      "epoch": 8.477533773730881,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 35612
    },
    {
      "epoch": 8.477771826459561,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 35613
    },
    {
      "epoch": 8.47800987918824,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7716,
      "step": 35614
    },
    {
      "epoch": 8.47824793191692,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 35615
    },
    {
      "epoch": 8.4784859846456,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 35616
    },
    {
      "epoch": 8.478724037374278,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2169,
      "step": 35617
    },
    {
      "epoch": 8.478962090102957,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1441,
      "step": 35618
    },
    {
      "epoch": 8.479200142831637,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.869,
      "step": 35619
    },
    {
      "epoch": 8.479438195560316,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0638,
      "step": 35620
    },
    {
      "epoch": 8.479676248288996,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 35621
    },
    {
      "epoch": 8.479914301017676,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9544,
      "step": 35622
    },
    {
      "epoch": 8.480152353746355,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 35623
    },
    {
      "epoch": 8.480390406475035,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 35624
    },
    {
      "epoch": 8.480628459203714,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 35625
    },
    {
      "epoch": 8.480866511932392,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8357,
      "step": 35626
    },
    {
      "epoch": 8.481104564661072,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1239,
      "step": 35627
    },
    {
      "epoch": 8.481342617389751,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 35628
    },
    {
      "epoch": 8.481580670118431,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 35629
    },
    {
      "epoch": 8.48181872284711,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7682,
      "step": 35630
    },
    {
      "epoch": 8.48205677557579,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 35631
    },
    {
      "epoch": 8.48229482830447,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 35632
    },
    {
      "epoch": 8.48253288103315,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 35633
    },
    {
      "epoch": 8.482770933761827,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 35634
    },
    {
      "epoch": 8.483008986490507,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 35635
    },
    {
      "epoch": 8.483247039219187,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2997,
      "step": 35636
    },
    {
      "epoch": 8.483485091947866,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 35637
    },
    {
      "epoch": 8.483723144676546,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 35638
    },
    {
      "epoch": 8.483961197405225,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 35639
    },
    {
      "epoch": 8.484199250133905,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 35640
    },
    {
      "epoch": 8.484437302862585,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 35641
    },
    {
      "epoch": 8.484675355591264,
      "grad_norm": 7.09375,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 35642
    },
    {
      "epoch": 8.484913408319942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7236,
      "step": 35643
    },
    {
      "epoch": 8.485151461048622,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 35644
    },
    {
      "epoch": 8.485389513777301,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2743,
      "step": 35645
    },
    {
      "epoch": 8.48562756650598,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0686,
      "step": 35646
    },
    {
      "epoch": 8.48586561923466,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1298,
      "step": 35647
    },
    {
      "epoch": 8.48610367196334,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 35648
    },
    {
      "epoch": 8.48634172469202,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 35649
    },
    {
      "epoch": 8.4865797774207,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0444,
      "step": 35650
    },
    {
      "epoch": 8.486817830149379,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 35651
    },
    {
      "epoch": 8.487055882878057,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 35652
    },
    {
      "epoch": 8.487293935606736,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 35653
    },
    {
      "epoch": 8.487531988335416,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 35654
    },
    {
      "epoch": 8.487770041064095,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3263,
      "step": 35655
    },
    {
      "epoch": 8.488008093792775,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 35656
    },
    {
      "epoch": 8.488246146521455,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8896,
      "step": 35657
    },
    {
      "epoch": 8.488484199250134,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 35658
    },
    {
      "epoch": 8.488722251978814,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 35659
    },
    {
      "epoch": 8.488960304707494,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9342,
      "step": 35660
    },
    {
      "epoch": 8.489198357436171,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 35661
    },
    {
      "epoch": 8.489436410164851,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 35662
    },
    {
      "epoch": 8.48967446289353,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 35663
    },
    {
      "epoch": 8.48991251562221,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 35664
    },
    {
      "epoch": 8.49015056835089,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 35665
    },
    {
      "epoch": 8.49038862107957,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7564,
      "step": 35666
    },
    {
      "epoch": 8.490626673808249,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 35667
    },
    {
      "epoch": 8.490864726536929,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 35668
    },
    {
      "epoch": 8.491102779265608,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1022,
      "step": 35669
    },
    {
      "epoch": 8.491340831994286,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 35670
    },
    {
      "epoch": 8.491578884722966,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 35671
    },
    {
      "epoch": 8.491816937451645,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 35672
    },
    {
      "epoch": 8.492054990180325,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 35673
    },
    {
      "epoch": 8.492293042909004,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 35674
    },
    {
      "epoch": 8.492531095637684,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 35675
    },
    {
      "epoch": 8.492769148366364,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1388,
      "step": 35676
    },
    {
      "epoch": 8.493007201095043,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 35677
    },
    {
      "epoch": 8.493245253823723,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7286,
      "step": 35678
    },
    {
      "epoch": 8.4934833065524,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9886,
      "step": 35679
    },
    {
      "epoch": 8.49372135928108,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2441,
      "step": 35680
    },
    {
      "epoch": 8.49395941200976,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0755,
      "step": 35681
    },
    {
      "epoch": 8.49419746473844,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 35682
    },
    {
      "epoch": 8.49443551746712,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.9077,
      "step": 35683
    },
    {
      "epoch": 8.494673570195799,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.3484,
      "step": 35684
    },
    {
      "epoch": 8.494911622924478,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2378,
      "step": 35685
    },
    {
      "epoch": 8.495149675653158,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1174,
      "step": 35686
    },
    {
      "epoch": 8.495387728381836,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0813,
      "step": 35687
    },
    {
      "epoch": 8.495625781110515,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8362,
      "step": 35688
    },
    {
      "epoch": 8.495863833839195,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 35689
    },
    {
      "epoch": 8.496101886567875,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 35690
    },
    {
      "epoch": 8.496339939296554,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1867,
      "step": 35691
    },
    {
      "epoch": 8.496577992025234,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1412,
      "step": 35692
    },
    {
      "epoch": 8.496816044753913,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1838,
      "step": 35693
    },
    {
      "epoch": 8.497054097482593,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0305,
      "step": 35694
    },
    {
      "epoch": 8.497292150211273,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 35695
    },
    {
      "epoch": 8.49753020293995,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1611,
      "step": 35696
    },
    {
      "epoch": 8.49776825566863,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0632,
      "step": 35697
    },
    {
      "epoch": 8.49800630839731,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 35698
    },
    {
      "epoch": 8.49824436112599,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0412,
      "step": 35699
    },
    {
      "epoch": 8.498482413854669,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 35700
    },
    {
      "epoch": 8.498482413854669,
      "eval_runtime": 0.5164,
      "eval_samples_per_second": 1.936,
      "eval_steps_per_second": 1.936,
      "step": 35700
    },
    {
      "epoch": 8.498720466583348,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9271,
      "step": 35701
    },
    {
      "epoch": 8.498958519312028,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 35702
    },
    {
      "epoch": 8.499196572040708,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 35703
    },
    {
      "epoch": 8.499434624769387,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 35704
    },
    {
      "epoch": 8.499672677498065,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2118,
      "step": 35705
    },
    {
      "epoch": 8.499910730226745,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 35706
    },
    {
      "epoch": 8.500148782955424,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.1222,
      "step": 35707
    },
    {
      "epoch": 8.500386835684104,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2799,
      "step": 35708
    },
    {
      "epoch": 8.500624888412784,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 35709
    },
    {
      "epoch": 8.500862941141463,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 35710
    },
    {
      "epoch": 8.501100993870143,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 35711
    },
    {
      "epoch": 8.501339046598822,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 35712
    },
    {
      "epoch": 8.501577099327502,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 35713
    },
    {
      "epoch": 8.50181515205618,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 35714
    },
    {
      "epoch": 8.50205320478486,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1465,
      "step": 35715
    },
    {
      "epoch": 8.502291257513539,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.063,
      "step": 35716
    },
    {
      "epoch": 8.502529310242219,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 35717
    },
    {
      "epoch": 8.502767362970898,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 35718
    },
    {
      "epoch": 8.503005415699578,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3686,
      "step": 35719
    },
    {
      "epoch": 8.503243468428257,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 35720
    },
    {
      "epoch": 8.503481521156937,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 35721
    },
    {
      "epoch": 8.503719573885615,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8129,
      "step": 35722
    },
    {
      "epoch": 8.503957626614294,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 35723
    },
    {
      "epoch": 8.504195679342974,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2708,
      "step": 35724
    },
    {
      "epoch": 8.504433732071654,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8474,
      "step": 35725
    },
    {
      "epoch": 8.504671784800333,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 35726
    },
    {
      "epoch": 8.504909837529013,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.806,
      "step": 35727
    },
    {
      "epoch": 8.505147890257692,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.811,
      "step": 35728
    },
    {
      "epoch": 8.505385942986372,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 35729
    },
    {
      "epoch": 8.505623995715052,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7778,
      "step": 35730
    },
    {
      "epoch": 8.50586204844373,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 35731
    },
    {
      "epoch": 8.506100101172409,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 35732
    },
    {
      "epoch": 8.506338153901089,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4305,
      "step": 35733
    },
    {
      "epoch": 8.506576206629768,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2494,
      "step": 35734
    },
    {
      "epoch": 8.506814259358448,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 35735
    },
    {
      "epoch": 8.507052312087128,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 35736
    },
    {
      "epoch": 8.507290364815807,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1548,
      "step": 35737
    },
    {
      "epoch": 8.507528417544487,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 35738
    },
    {
      "epoch": 8.507766470273165,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8101,
      "step": 35739
    },
    {
      "epoch": 8.508004523001844,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 35740
    },
    {
      "epoch": 8.508242575730524,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9356,
      "step": 35741
    },
    {
      "epoch": 8.508480628459203,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 35742
    },
    {
      "epoch": 8.508718681187883,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9055,
      "step": 35743
    },
    {
      "epoch": 8.508956733916563,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 35744
    },
    {
      "epoch": 8.509194786645242,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 35745
    },
    {
      "epoch": 8.509432839373922,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 35746
    },
    {
      "epoch": 8.509670892102601,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 35747
    },
    {
      "epoch": 8.509908944831281,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 35748
    },
    {
      "epoch": 8.510146997559959,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3313,
      "step": 35749
    },
    {
      "epoch": 8.510385050288638,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 35750
    },
    {
      "epoch": 8.510623103017318,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.024,
      "step": 35751
    },
    {
      "epoch": 8.510861155745998,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7691,
      "step": 35752
    },
    {
      "epoch": 8.511099208474677,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 35753
    },
    {
      "epoch": 8.511337261203357,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8046,
      "step": 35754
    },
    {
      "epoch": 8.511575313932036,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 35755
    },
    {
      "epoch": 8.511813366660716,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 35756
    },
    {
      "epoch": 8.512051419389394,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.099,
      "step": 35757
    },
    {
      "epoch": 8.512289472118074,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 35758
    },
    {
      "epoch": 8.512527524846753,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 35759
    },
    {
      "epoch": 8.512765577575433,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9491,
      "step": 35760
    },
    {
      "epoch": 8.513003630304112,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 35761
    },
    {
      "epoch": 8.513241683032792,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0098,
      "step": 35762
    },
    {
      "epoch": 8.513479735761472,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.737,
      "step": 35763
    },
    {
      "epoch": 8.513717788490151,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0831,
      "step": 35764
    },
    {
      "epoch": 8.51395584121883,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 35765
    },
    {
      "epoch": 8.514193893947509,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9343,
      "step": 35766
    },
    {
      "epoch": 8.514431946676188,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2078,
      "step": 35767
    },
    {
      "epoch": 8.514669999404868,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 35768
    },
    {
      "epoch": 8.514908052133547,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 35769
    },
    {
      "epoch": 8.515146104862227,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 35770
    },
    {
      "epoch": 8.515384157590907,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 35771
    },
    {
      "epoch": 8.515622210319586,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9897,
      "step": 35772
    },
    {
      "epoch": 8.515860263048266,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1481,
      "step": 35773
    },
    {
      "epoch": 8.516098315776944,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 35774
    },
    {
      "epoch": 8.516336368505623,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 35775
    },
    {
      "epoch": 8.516574421234303,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 35776
    },
    {
      "epoch": 8.516812473962982,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3177,
      "step": 35777
    },
    {
      "epoch": 8.517050526691662,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 35778
    },
    {
      "epoch": 8.517288579420342,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6585,
      "step": 35779
    },
    {
      "epoch": 8.517526632149021,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0189,
      "step": 35780
    },
    {
      "epoch": 8.5177646848777,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 35781
    },
    {
      "epoch": 8.51800273760638,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 35782
    },
    {
      "epoch": 8.51824079033506,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 35783
    },
    {
      "epoch": 8.518478843063738,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1279,
      "step": 35784
    },
    {
      "epoch": 8.518716895792418,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.054,
      "step": 35785
    },
    {
      "epoch": 8.518954948521097,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 35786
    },
    {
      "epoch": 8.519193001249777,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 35787
    },
    {
      "epoch": 8.519431053978456,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 35788
    },
    {
      "epoch": 8.519669106707136,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 35789
    },
    {
      "epoch": 8.519907159435816,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 35790
    },
    {
      "epoch": 8.520145212164495,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 35791
    },
    {
      "epoch": 8.520383264893173,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.212,
      "step": 35792
    },
    {
      "epoch": 8.520621317621853,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.299,
      "step": 35793
    },
    {
      "epoch": 8.520859370350532,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9406,
      "step": 35794
    },
    {
      "epoch": 8.521097423079212,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 35795
    },
    {
      "epoch": 8.521335475807891,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 35796
    },
    {
      "epoch": 8.521573528536571,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 35797
    },
    {
      "epoch": 8.52181158126525,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0787,
      "step": 35798
    },
    {
      "epoch": 8.52204963399393,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 35799
    },
    {
      "epoch": 8.52228768672261,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 35800
    },
    {
      "epoch": 8.52228768672261,
      "eval_runtime": 0.5195,
      "eval_samples_per_second": 1.925,
      "eval_steps_per_second": 1.925,
      "step": 35800
    },
    {
      "epoch": 8.522525739451288,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8923,
      "step": 35801
    },
    {
      "epoch": 8.522763792179967,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 35802
    },
    {
      "epoch": 8.523001844908647,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0481,
      "step": 35803
    },
    {
      "epoch": 8.523239897637326,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 35804
    },
    {
      "epoch": 8.523477950366006,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9692,
      "step": 35805
    },
    {
      "epoch": 8.523716003094686,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7444,
      "step": 35806
    },
    {
      "epoch": 8.523954055823365,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.328,
      "step": 35807
    },
    {
      "epoch": 8.524192108552045,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 35808
    },
    {
      "epoch": 8.524430161280725,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.1922,
      "step": 35809
    },
    {
      "epoch": 8.524668214009402,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 35810
    },
    {
      "epoch": 8.524906266738082,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9097,
      "step": 35811
    },
    {
      "epoch": 8.525144319466762,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 35812
    },
    {
      "epoch": 8.525382372195441,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 35813
    },
    {
      "epoch": 8.52562042492412,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 35814
    },
    {
      "epoch": 8.5258584776528,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8458,
      "step": 35815
    },
    {
      "epoch": 8.52609653038148,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7608,
      "step": 35816
    },
    {
      "epoch": 8.52633458311016,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 35817
    },
    {
      "epoch": 8.52657263583884,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 35818
    },
    {
      "epoch": 8.526810688567517,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 35819
    },
    {
      "epoch": 8.527048741296197,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0066,
      "step": 35820
    },
    {
      "epoch": 8.527286794024876,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 35821
    },
    {
      "epoch": 8.527524846753556,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 35822
    },
    {
      "epoch": 8.527762899482235,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0996,
      "step": 35823
    },
    {
      "epoch": 8.528000952210915,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 35824
    },
    {
      "epoch": 8.528239004939595,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 35825
    },
    {
      "epoch": 8.528477057668274,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 35826
    },
    {
      "epoch": 8.528715110396952,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 35827
    },
    {
      "epoch": 8.528953163125632,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.3379,
      "step": 35828
    },
    {
      "epoch": 8.529191215854311,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3042,
      "step": 35829
    },
    {
      "epoch": 8.52942926858299,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 35830
    },
    {
      "epoch": 8.52966732131167,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1663,
      "step": 35831
    },
    {
      "epoch": 8.52990537404035,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 35832
    },
    {
      "epoch": 8.53014342676903,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 35833
    },
    {
      "epoch": 8.53038147949771,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9555,
      "step": 35834
    },
    {
      "epoch": 8.530619532226389,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 35835
    },
    {
      "epoch": 8.530857584955067,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 35836
    },
    {
      "epoch": 8.531095637683746,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 35837
    },
    {
      "epoch": 8.531333690412426,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 35838
    },
    {
      "epoch": 8.531571743141106,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 35839
    },
    {
      "epoch": 8.531809795869785,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 35840
    },
    {
      "epoch": 8.532047848598465,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 35841
    },
    {
      "epoch": 8.532285901327144,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 35842
    },
    {
      "epoch": 8.532523954055824,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.627,
      "step": 35843
    },
    {
      "epoch": 8.532762006784504,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9572,
      "step": 35844
    },
    {
      "epoch": 8.533000059513181,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.7102,
      "step": 35845
    },
    {
      "epoch": 8.533238112241861,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7692,
      "step": 35846
    },
    {
      "epoch": 8.53347616497054,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 35847
    },
    {
      "epoch": 8.53371421769922,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 35848
    },
    {
      "epoch": 8.5339522704279,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 35849
    },
    {
      "epoch": 8.53419032315658,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 35850
    },
    {
      "epoch": 8.534428375885259,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 35851
    },
    {
      "epoch": 8.534666428613939,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9533,
      "step": 35852
    },
    {
      "epoch": 8.534904481342618,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 35853
    },
    {
      "epoch": 8.535142534071296,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2218,
      "step": 35854
    },
    {
      "epoch": 8.535380586799976,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 35855
    },
    {
      "epoch": 8.535618639528655,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 35856
    },
    {
      "epoch": 8.535856692257335,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1507,
      "step": 35857
    },
    {
      "epoch": 8.536094744986015,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.451,
      "step": 35858
    },
    {
      "epoch": 8.536332797714694,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2312,
      "step": 35859
    },
    {
      "epoch": 8.536570850443374,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0146,
      "step": 35860
    },
    {
      "epoch": 8.536808903172053,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 35861
    },
    {
      "epoch": 8.537046955900731,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 35862
    },
    {
      "epoch": 8.53728500862941,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1834,
      "step": 35863
    },
    {
      "epoch": 8.53752306135809,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.881,
      "step": 35864
    },
    {
      "epoch": 8.53776111408677,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0673,
      "step": 35865
    },
    {
      "epoch": 8.53799916681545,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 35866
    },
    {
      "epoch": 8.53823721954413,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 35867
    },
    {
      "epoch": 8.538475272272809,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8056,
      "step": 35868
    },
    {
      "epoch": 8.538713325001488,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 35869
    },
    {
      "epoch": 8.538951377730168,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 35870
    },
    {
      "epoch": 8.539189430458846,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 35871
    },
    {
      "epoch": 8.539427483187525,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7579,
      "step": 35872
    },
    {
      "epoch": 8.539665535916205,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.2245,
      "step": 35873
    },
    {
      "epoch": 8.539903588644885,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.3457,
      "step": 35874
    },
    {
      "epoch": 8.540141641373564,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3211,
      "step": 35875
    },
    {
      "epoch": 8.540379694102244,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 35876
    },
    {
      "epoch": 8.540617746830923,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 35877
    },
    {
      "epoch": 8.540855799559603,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 35878
    },
    {
      "epoch": 8.541093852288283,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1247,
      "step": 35879
    },
    {
      "epoch": 8.54133190501696,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 35880
    },
    {
      "epoch": 8.54156995774564,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 35881
    },
    {
      "epoch": 8.54180801047432,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 35882
    },
    {
      "epoch": 8.542046063203,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 35883
    },
    {
      "epoch": 8.542284115931679,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 35884
    },
    {
      "epoch": 8.542522168660359,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9031,
      "step": 35885
    },
    {
      "epoch": 8.542760221389038,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 35886
    },
    {
      "epoch": 8.542998274117718,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 35887
    },
    {
      "epoch": 8.543236326846397,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 35888
    },
    {
      "epoch": 8.543474379575075,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 35889
    },
    {
      "epoch": 8.543712432303755,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1225,
      "step": 35890
    },
    {
      "epoch": 8.543950485032434,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 35891
    },
    {
      "epoch": 8.544188537761114,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7877,
      "step": 35892
    },
    {
      "epoch": 8.544426590489794,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1042,
      "step": 35893
    },
    {
      "epoch": 8.544664643218473,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2884,
      "step": 35894
    },
    {
      "epoch": 8.544902695947153,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 35895
    },
    {
      "epoch": 8.545140748675832,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 35896
    },
    {
      "epoch": 8.54537880140451,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 35897
    },
    {
      "epoch": 8.54561685413319,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 35898
    },
    {
      "epoch": 8.54585490686187,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 35899
    },
    {
      "epoch": 8.546092959590549,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 35900
    },
    {
      "epoch": 8.546092959590549,
      "eval_runtime": 0.5237,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 1.91,
      "step": 35900
    },
    {
      "epoch": 8.546331012319229,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 35901
    },
    {
      "epoch": 8.546569065047908,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 35902
    },
    {
      "epoch": 8.546807117776588,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 35903
    },
    {
      "epoch": 8.547045170505267,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 35904
    },
    {
      "epoch": 8.547283223233947,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 35905
    },
    {
      "epoch": 8.547521275962625,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7737,
      "step": 35906
    },
    {
      "epoch": 8.547759328691305,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 35907
    },
    {
      "epoch": 8.547997381419984,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.3827,
      "step": 35908
    },
    {
      "epoch": 8.548235434148664,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7408,
      "step": 35909
    },
    {
      "epoch": 8.548473486877343,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 35910
    },
    {
      "epoch": 8.548711539606023,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 35911
    },
    {
      "epoch": 8.548949592334703,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 35912
    },
    {
      "epoch": 8.549187645063382,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 35913
    },
    {
      "epoch": 8.549425697792062,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8729,
      "step": 35914
    },
    {
      "epoch": 8.54966375052074,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 35915
    },
    {
      "epoch": 8.54990180324942,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 35916
    },
    {
      "epoch": 8.550139855978099,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8536,
      "step": 35917
    },
    {
      "epoch": 8.550377908706778,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 35918
    },
    {
      "epoch": 8.550615961435458,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 35919
    },
    {
      "epoch": 8.550854014164138,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 35920
    },
    {
      "epoch": 8.551092066892817,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 35921
    },
    {
      "epoch": 8.551330119621497,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7684,
      "step": 35922
    },
    {
      "epoch": 8.551568172350176,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.36,
      "step": 35923
    },
    {
      "epoch": 8.551806225078854,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 35924
    },
    {
      "epoch": 8.552044277807534,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 35925
    },
    {
      "epoch": 8.552282330536213,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8964,
      "step": 35926
    },
    {
      "epoch": 8.552520383264893,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 35927
    },
    {
      "epoch": 8.552758435993573,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 35928
    },
    {
      "epoch": 8.552996488722252,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 35929
    },
    {
      "epoch": 8.553234541450932,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 35930
    },
    {
      "epoch": 8.553472594179611,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0759,
      "step": 35931
    },
    {
      "epoch": 8.55371064690829,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.418,
      "step": 35932
    },
    {
      "epoch": 8.553948699636969,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 35933
    },
    {
      "epoch": 8.554186752365649,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9136,
      "step": 35934
    },
    {
      "epoch": 8.554424805094328,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.874,
      "step": 35935
    },
    {
      "epoch": 8.554662857823008,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 35936
    },
    {
      "epoch": 8.554900910551687,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2615,
      "step": 35937
    },
    {
      "epoch": 8.555138963280367,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 35938
    },
    {
      "epoch": 8.555377016009047,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 35939
    },
    {
      "epoch": 8.555615068737726,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 35940
    },
    {
      "epoch": 8.555853121466404,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 35941
    },
    {
      "epoch": 8.556091174195084,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0905,
      "step": 35942
    },
    {
      "epoch": 8.556329226923763,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9981,
      "step": 35943
    },
    {
      "epoch": 8.556567279652443,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 35944
    },
    {
      "epoch": 8.556805332381122,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 35945
    },
    {
      "epoch": 8.557043385109802,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 35946
    },
    {
      "epoch": 8.557281437838482,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 35947
    },
    {
      "epoch": 8.557519490567161,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 35948
    },
    {
      "epoch": 8.55775754329584,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 35949
    },
    {
      "epoch": 8.557995596024519,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8519,
      "step": 35950
    },
    {
      "epoch": 8.558233648753198,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3738,
      "step": 35951
    },
    {
      "epoch": 8.558471701481878,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 35952
    },
    {
      "epoch": 8.558709754210557,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 35953
    },
    {
      "epoch": 8.558947806939237,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 35954
    },
    {
      "epoch": 8.559185859667917,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 35955
    },
    {
      "epoch": 8.559423912396596,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 35956
    },
    {
      "epoch": 8.559661965125276,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 35957
    },
    {
      "epoch": 8.559900017853955,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 35958
    },
    {
      "epoch": 8.560138070582633,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 35959
    },
    {
      "epoch": 8.560376123311313,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 35960
    },
    {
      "epoch": 8.560614176039993,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.3156,
      "step": 35961
    },
    {
      "epoch": 8.560852228768672,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.004,
      "step": 35962
    },
    {
      "epoch": 8.561090281497352,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 35963
    },
    {
      "epoch": 8.561328334226031,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 35964
    },
    {
      "epoch": 8.561566386954711,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.3561,
      "step": 35965
    },
    {
      "epoch": 8.56180443968339,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.2519,
      "step": 35966
    },
    {
      "epoch": 8.562042492412068,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.2648,
      "step": 35967
    },
    {
      "epoch": 8.562280545140748,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9636,
      "step": 35968
    },
    {
      "epoch": 8.562518597869428,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3992,
      "step": 35969
    },
    {
      "epoch": 8.562756650598107,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 35970
    },
    {
      "epoch": 8.562994703326787,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 35971
    },
    {
      "epoch": 8.563232756055466,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0811,
      "step": 35972
    },
    {
      "epoch": 8.563470808784146,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 35973
    },
    {
      "epoch": 8.563708861512826,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 35974
    },
    {
      "epoch": 8.563946914241505,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7576,
      "step": 35975
    },
    {
      "epoch": 8.564184966970185,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 35976
    },
    {
      "epoch": 8.564423019698863,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 35977
    },
    {
      "epoch": 8.564661072427542,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 35978
    },
    {
      "epoch": 8.564899125156222,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9364,
      "step": 35979
    },
    {
      "epoch": 8.565137177884901,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.043,
      "step": 35980
    },
    {
      "epoch": 8.565375230613581,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 35981
    },
    {
      "epoch": 8.56561328334226,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 35982
    },
    {
      "epoch": 8.56585133607094,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 35983
    },
    {
      "epoch": 8.56608938879962,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 35984
    },
    {
      "epoch": 8.566327441528298,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7744,
      "step": 35985
    },
    {
      "epoch": 8.566565494256977,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 35986
    },
    {
      "epoch": 8.566803546985657,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 35987
    },
    {
      "epoch": 8.567041599714337,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 35988
    },
    {
      "epoch": 8.567279652443016,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 35989
    },
    {
      "epoch": 8.567517705171696,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 35990
    },
    {
      "epoch": 8.567755757900375,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 35991
    },
    {
      "epoch": 8.567993810629055,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 35992
    },
    {
      "epoch": 8.568231863357735,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.6549,
      "step": 35993
    },
    {
      "epoch": 8.568469916086412,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 35994
    },
    {
      "epoch": 8.568707968815092,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 35995
    },
    {
      "epoch": 8.568946021543772,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 35996
    },
    {
      "epoch": 8.569184074272451,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0708,
      "step": 35997
    },
    {
      "epoch": 8.56942212700113,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 35998
    },
    {
      "epoch": 8.56966017972981,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 35999
    },
    {
      "epoch": 8.56989823245849,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2733,
      "step": 36000
    },
    {
      "epoch": 8.56989823245849,
      "eval_runtime": 0.5139,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 36000
    },
    {
      "epoch": 8.57013628518717,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 36001
    },
    {
      "epoch": 8.570374337915847,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9216,
      "step": 36002
    },
    {
      "epoch": 8.570612390644527,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2903,
      "step": 36003
    },
    {
      "epoch": 8.570850443373207,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 36004
    },
    {
      "epoch": 8.571088496101886,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7283,
      "step": 36005
    },
    {
      "epoch": 8.571326548830566,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 36006
    },
    {
      "epoch": 8.571564601559245,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0133,
      "step": 36007
    },
    {
      "epoch": 8.571802654287925,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8344,
      "step": 36008
    },
    {
      "epoch": 8.572040707016605,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 36009
    },
    {
      "epoch": 8.572278759745284,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 36010
    },
    {
      "epoch": 8.572516812473964,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 36011
    },
    {
      "epoch": 8.572754865202642,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 36012
    },
    {
      "epoch": 8.572992917931321,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0418,
      "step": 36013
    },
    {
      "epoch": 8.573230970660001,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.346,
      "step": 36014
    },
    {
      "epoch": 8.57346902338868,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 36015
    },
    {
      "epoch": 8.57370707611736,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 36016
    },
    {
      "epoch": 8.57394512884604,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0182,
      "step": 36017
    },
    {
      "epoch": 8.57418318157472,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 36018
    },
    {
      "epoch": 8.574421234303399,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.4351,
      "step": 36019
    },
    {
      "epoch": 8.574659287032077,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 36020
    },
    {
      "epoch": 8.574897339760756,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 36021
    },
    {
      "epoch": 8.575135392489436,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8656,
      "step": 36022
    },
    {
      "epoch": 8.575373445218116,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 36023
    },
    {
      "epoch": 8.575611497946795,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 36024
    },
    {
      "epoch": 8.575849550675475,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 36025
    },
    {
      "epoch": 8.576087603404154,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 36026
    },
    {
      "epoch": 8.576325656132834,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 36027
    },
    {
      "epoch": 8.576563708861514,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1835,
      "step": 36028
    },
    {
      "epoch": 8.576801761590191,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3844,
      "step": 36029
    },
    {
      "epoch": 8.577039814318871,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 36030
    },
    {
      "epoch": 8.57727786704755,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8437,
      "step": 36031
    },
    {
      "epoch": 8.57751591977623,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 36032
    },
    {
      "epoch": 8.57775397250491,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1381,
      "step": 36033
    },
    {
      "epoch": 8.57799202523359,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 36034
    },
    {
      "epoch": 8.578230077962269,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3403,
      "step": 36035
    },
    {
      "epoch": 8.578468130690949,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.186,
      "step": 36036
    },
    {
      "epoch": 8.578706183419627,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 36037
    },
    {
      "epoch": 8.578944236148306,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 36038
    },
    {
      "epoch": 8.579182288876986,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 36039
    },
    {
      "epoch": 8.579420341605665,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.7507,
      "step": 36040
    },
    {
      "epoch": 8.579658394334345,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 36041
    },
    {
      "epoch": 8.579896447063025,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 36042
    },
    {
      "epoch": 8.580134499791704,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1336,
      "step": 36043
    },
    {
      "epoch": 8.580372552520384,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 36044
    },
    {
      "epoch": 8.580610605249063,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0927,
      "step": 36045
    },
    {
      "epoch": 8.580848657977743,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 36046
    },
    {
      "epoch": 8.58108671070642,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 36047
    },
    {
      "epoch": 8.5813247634351,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 36048
    },
    {
      "epoch": 8.58156281616378,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 36049
    },
    {
      "epoch": 8.58180086889246,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 1.6905,
      "step": 36050
    },
    {
      "epoch": 8.58203892162114,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 36051
    },
    {
      "epoch": 8.582276974349819,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0679,
      "step": 36052
    },
    {
      "epoch": 8.582515027078498,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 36053
    },
    {
      "epoch": 8.582753079807178,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.033,
      "step": 36054
    },
    {
      "epoch": 8.582991132535856,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 36055
    },
    {
      "epoch": 8.583229185264535,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2291,
      "step": 36056
    },
    {
      "epoch": 8.583467237993215,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 36057
    },
    {
      "epoch": 8.583705290721895,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 36058
    },
    {
      "epoch": 8.583943343450574,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 36059
    },
    {
      "epoch": 8.584181396179254,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 36060
    },
    {
      "epoch": 8.584419448907934,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0709,
      "step": 36061
    },
    {
      "epoch": 8.584657501636613,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8553,
      "step": 36062
    },
    {
      "epoch": 8.584895554365293,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 36063
    },
    {
      "epoch": 8.58513360709397,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7646,
      "step": 36064
    },
    {
      "epoch": 8.58537165982265,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 36065
    },
    {
      "epoch": 8.58560971255133,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 36066
    },
    {
      "epoch": 8.58584776528001,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0531,
      "step": 36067
    },
    {
      "epoch": 8.586085818008689,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 36068
    },
    {
      "epoch": 8.586323870737369,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 36069
    },
    {
      "epoch": 8.586561923466048,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0094,
      "step": 36070
    },
    {
      "epoch": 8.586799976194728,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0943,
      "step": 36071
    },
    {
      "epoch": 8.587038028923406,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2613,
      "step": 36072
    },
    {
      "epoch": 8.587276081652085,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 36073
    },
    {
      "epoch": 8.587514134380765,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9268,
      "step": 36074
    },
    {
      "epoch": 8.587752187109444,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.2497,
      "step": 36075
    },
    {
      "epoch": 8.587990239838124,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0351,
      "step": 36076
    },
    {
      "epoch": 8.588228292566804,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 36077
    },
    {
      "epoch": 8.588466345295483,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9723,
      "step": 36078
    },
    {
      "epoch": 8.588704398024163,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1807,
      "step": 36079
    },
    {
      "epoch": 8.588942450752842,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 36080
    },
    {
      "epoch": 8.589180503481522,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 36081
    },
    {
      "epoch": 8.5894185562102,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 36082
    },
    {
      "epoch": 8.58965660893888,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 36083
    },
    {
      "epoch": 8.589894661667559,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 36084
    },
    {
      "epoch": 8.590132714396239,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 36085
    },
    {
      "epoch": 8.590370767124918,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7984,
      "step": 36086
    },
    {
      "epoch": 8.590608819853598,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.3205,
      "step": 36087
    },
    {
      "epoch": 8.590846872582278,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 36088
    },
    {
      "epoch": 8.591084925310957,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 36089
    },
    {
      "epoch": 8.591322978039635,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 36090
    },
    {
      "epoch": 8.591561030768315,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2537,
      "step": 36091
    },
    {
      "epoch": 8.591799083496994,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7811,
      "step": 36092
    },
    {
      "epoch": 8.592037136225674,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 36093
    },
    {
      "epoch": 8.592275188954353,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 36094
    },
    {
      "epoch": 8.592513241683033,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 36095
    },
    {
      "epoch": 8.592751294411713,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1887,
      "step": 36096
    },
    {
      "epoch": 8.592989347140392,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 36097
    },
    {
      "epoch": 8.593227399869072,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 36098
    },
    {
      "epoch": 8.59346545259775,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 36099
    },
    {
      "epoch": 8.59370350532643,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 36100
    },
    {
      "epoch": 8.59370350532643,
      "eval_runtime": 0.5175,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 36100
    },
    {
      "epoch": 8.593941558055109,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 36101
    },
    {
      "epoch": 8.594179610783788,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 36102
    },
    {
      "epoch": 8.594417663512468,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2933,
      "step": 36103
    },
    {
      "epoch": 8.594655716241148,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 36104
    },
    {
      "epoch": 8.594893768969827,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 36105
    },
    {
      "epoch": 8.595131821698507,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 36106
    },
    {
      "epoch": 8.595369874427185,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1443,
      "step": 36107
    },
    {
      "epoch": 8.595607927155864,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2299,
      "step": 36108
    },
    {
      "epoch": 8.595845979884544,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 36109
    },
    {
      "epoch": 8.596084032613224,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0577,
      "step": 36110
    },
    {
      "epoch": 8.596322085341903,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 36111
    },
    {
      "epoch": 8.596560138070583,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2135,
      "step": 36112
    },
    {
      "epoch": 8.596798190799262,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2871,
      "step": 36113
    },
    {
      "epoch": 8.597036243527942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 36114
    },
    {
      "epoch": 8.597274296256622,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 36115
    },
    {
      "epoch": 8.597512348985301,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8473,
      "step": 36116
    },
    {
      "epoch": 8.597750401713979,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 36117
    },
    {
      "epoch": 8.597988454442659,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 36118
    },
    {
      "epoch": 8.598226507171338,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 36119
    },
    {
      "epoch": 8.598464559900018,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.094,
      "step": 36120
    },
    {
      "epoch": 8.598702612628697,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 36121
    },
    {
      "epoch": 8.598940665357377,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 36122
    },
    {
      "epoch": 8.599178718086057,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 36123
    },
    {
      "epoch": 8.599416770814736,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2161,
      "step": 36124
    },
    {
      "epoch": 8.599654823543414,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9065,
      "step": 36125
    },
    {
      "epoch": 8.599892876272094,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.185,
      "step": 36126
    },
    {
      "epoch": 8.600130929000773,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 36127
    },
    {
      "epoch": 8.600368981729453,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9762,
      "step": 36128
    },
    {
      "epoch": 8.600607034458132,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 36129
    },
    {
      "epoch": 8.600845087186812,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 36130
    },
    {
      "epoch": 8.601083139915492,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 36131
    },
    {
      "epoch": 8.601321192644171,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8085,
      "step": 36132
    },
    {
      "epoch": 8.60155924537285,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 36133
    },
    {
      "epoch": 8.601797298101529,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8016,
      "step": 36134
    },
    {
      "epoch": 8.602035350830208,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 36135
    },
    {
      "epoch": 8.602273403558888,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 36136
    },
    {
      "epoch": 8.602511456287568,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8926,
      "step": 36137
    },
    {
      "epoch": 8.602749509016247,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3545,
      "step": 36138
    },
    {
      "epoch": 8.602987561744927,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 36139
    },
    {
      "epoch": 8.603225614473606,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3673,
      "step": 36140
    },
    {
      "epoch": 8.603463667202286,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0881,
      "step": 36141
    },
    {
      "epoch": 8.603701719930964,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 36142
    },
    {
      "epoch": 8.603939772659643,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.017,
      "step": 36143
    },
    {
      "epoch": 8.604177825388323,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 36144
    },
    {
      "epoch": 8.604415878117003,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 36145
    },
    {
      "epoch": 8.604653930845682,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 36146
    },
    {
      "epoch": 8.604891983574362,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 36147
    },
    {
      "epoch": 8.605130036303041,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 36148
    },
    {
      "epoch": 8.605368089031721,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1606,
      "step": 36149
    },
    {
      "epoch": 8.6056061417604,
      "grad_norm": 8.125,
      "learning_rate": 3e-05,
      "loss": 2.1547,
      "step": 36150
    },
    {
      "epoch": 8.60584419448908,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2467,
      "step": 36151
    },
    {
      "epoch": 8.606082247217758,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 36152
    },
    {
      "epoch": 8.606320299946438,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 36153
    },
    {
      "epoch": 8.606558352675117,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3146,
      "step": 36154
    },
    {
      "epoch": 8.606796405403797,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 36155
    },
    {
      "epoch": 8.607034458132476,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0214,
      "step": 36156
    },
    {
      "epoch": 8.607272510861156,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 36157
    },
    {
      "epoch": 8.607510563589836,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9952,
      "step": 36158
    },
    {
      "epoch": 8.607748616318515,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8651,
      "step": 36159
    },
    {
      "epoch": 8.607986669047193,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1295,
      "step": 36160
    },
    {
      "epoch": 8.608224721775873,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1009,
      "step": 36161
    },
    {
      "epoch": 8.608462774504552,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 36162
    },
    {
      "epoch": 8.608700827233232,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 36163
    },
    {
      "epoch": 8.608938879961912,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 36164
    },
    {
      "epoch": 8.609176932690591,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.315,
      "step": 36165
    },
    {
      "epoch": 8.60941498541927,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 36166
    },
    {
      "epoch": 8.60965303814795,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 36167
    },
    {
      "epoch": 8.60989109087663,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0513,
      "step": 36168
    },
    {
      "epoch": 8.610129143605308,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 36169
    },
    {
      "epoch": 8.610367196333987,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 36170
    },
    {
      "epoch": 8.610605249062667,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1053,
      "step": 36171
    },
    {
      "epoch": 8.610843301791347,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 36172
    },
    {
      "epoch": 8.611081354520026,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.109,
      "step": 36173
    },
    {
      "epoch": 8.611319407248706,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 36174
    },
    {
      "epoch": 8.611557459977385,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2112,
      "step": 36175
    },
    {
      "epoch": 8.611795512706065,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 36176
    },
    {
      "epoch": 8.612033565434743,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 36177
    },
    {
      "epoch": 8.612271618163422,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 36178
    },
    {
      "epoch": 8.612509670892102,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 36179
    },
    {
      "epoch": 8.612747723620782,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 36180
    },
    {
      "epoch": 8.612985776349461,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8752,
      "step": 36181
    },
    {
      "epoch": 8.61322382907814,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8221,
      "step": 36182
    },
    {
      "epoch": 8.61346188180682,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 36183
    },
    {
      "epoch": 8.6136999345355,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 36184
    },
    {
      "epoch": 8.61393798726418,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7958,
      "step": 36185
    },
    {
      "epoch": 8.61417603999286,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 36186
    },
    {
      "epoch": 8.614414092721537,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 36187
    },
    {
      "epoch": 8.614652145450217,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 36188
    },
    {
      "epoch": 8.614890198178896,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2524,
      "step": 36189
    },
    {
      "epoch": 8.615128250907576,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 36190
    },
    {
      "epoch": 8.615366303636256,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9331,
      "step": 36191
    },
    {
      "epoch": 8.615604356364935,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 36192
    },
    {
      "epoch": 8.615842409093615,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 36193
    },
    {
      "epoch": 8.616080461822294,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 36194
    },
    {
      "epoch": 8.616318514550972,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 36195
    },
    {
      "epoch": 8.616556567279652,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8585,
      "step": 36196
    },
    {
      "epoch": 8.616794620008331,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 36197
    },
    {
      "epoch": 8.617032672737011,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 36198
    },
    {
      "epoch": 8.61727072546569,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 36199
    },
    {
      "epoch": 8.61750877819437,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 36200
    },
    {
      "epoch": 8.61750877819437,
      "eval_runtime": 0.5343,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 1.872,
      "step": 36200
    },
    {
      "epoch": 8.61774683092305,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0225,
      "step": 36201
    },
    {
      "epoch": 8.61798488365173,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 36202
    },
    {
      "epoch": 8.618222936380409,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.6183,
      "step": 36203
    },
    {
      "epoch": 8.618460989109087,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 36204
    },
    {
      "epoch": 8.618699041837766,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 36205
    },
    {
      "epoch": 8.618937094566446,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 36206
    },
    {
      "epoch": 8.619175147295126,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3799,
      "step": 36207
    },
    {
      "epoch": 8.619413200023805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9142,
      "step": 36208
    },
    {
      "epoch": 8.619651252752485,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8041,
      "step": 36209
    },
    {
      "epoch": 8.619889305481164,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1647,
      "step": 36210
    },
    {
      "epoch": 8.620127358209844,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 36211
    },
    {
      "epoch": 8.620365410938524,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 36212
    },
    {
      "epoch": 8.620603463667202,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.6292,
      "step": 36213
    },
    {
      "epoch": 8.620841516395881,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 36214
    },
    {
      "epoch": 8.62107956912456,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3976,
      "step": 36215
    },
    {
      "epoch": 8.62131762185324,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1198,
      "step": 36216
    },
    {
      "epoch": 8.62155567458192,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.253,
      "step": 36217
    },
    {
      "epoch": 8.6217937273106,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 36218
    },
    {
      "epoch": 8.62203178003928,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.7786,
      "step": 36219
    },
    {
      "epoch": 8.622269832767959,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 36220
    },
    {
      "epoch": 8.622507885496638,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 36221
    },
    {
      "epoch": 8.622745938225316,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 36222
    },
    {
      "epoch": 8.622983990953996,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0076,
      "step": 36223
    },
    {
      "epoch": 8.623222043682675,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 36224
    },
    {
      "epoch": 8.623460096411355,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1429,
      "step": 36225
    },
    {
      "epoch": 8.623698149140035,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 36226
    },
    {
      "epoch": 8.623936201868714,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 36227
    },
    {
      "epoch": 8.624174254597394,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 36228
    },
    {
      "epoch": 8.624412307326073,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 36229
    },
    {
      "epoch": 8.624650360054751,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 36230
    },
    {
      "epoch": 8.62488841278343,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 36231
    },
    {
      "epoch": 8.62512646551211,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2694,
      "step": 36232
    },
    {
      "epoch": 8.62536451824079,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 36233
    },
    {
      "epoch": 8.62560257096947,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2055,
      "step": 36234
    },
    {
      "epoch": 8.62584062369815,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 36235
    },
    {
      "epoch": 8.626078676426829,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1883,
      "step": 36236
    },
    {
      "epoch": 8.626316729155509,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1593,
      "step": 36237
    },
    {
      "epoch": 8.626554781884188,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 36238
    },
    {
      "epoch": 8.626792834612866,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.778,
      "step": 36239
    },
    {
      "epoch": 8.627030887341546,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 36240
    },
    {
      "epoch": 8.627268940070225,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.7057,
      "step": 36241
    },
    {
      "epoch": 8.627506992798905,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8448,
      "step": 36242
    },
    {
      "epoch": 8.627745045527584,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 36243
    },
    {
      "epoch": 8.627983098256264,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 36244
    },
    {
      "epoch": 8.628221150984944,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 36245
    },
    {
      "epoch": 8.628459203713623,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 36246
    },
    {
      "epoch": 8.628697256442303,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0639,
      "step": 36247
    },
    {
      "epoch": 8.62893530917098,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 36248
    },
    {
      "epoch": 8.62917336189966,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 36249
    },
    {
      "epoch": 8.62941141462834,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 36250
    },
    {
      "epoch": 8.62964946735702,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 36251
    },
    {
      "epoch": 8.629887520085699,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7901,
      "step": 36252
    },
    {
      "epoch": 8.630125572814379,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 36253
    },
    {
      "epoch": 8.630363625543058,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 36254
    },
    {
      "epoch": 8.630601678271738,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 36255
    },
    {
      "epoch": 8.630839731000417,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3449,
      "step": 36256
    },
    {
      "epoch": 8.631077783729095,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0749,
      "step": 36257
    },
    {
      "epoch": 8.631315836457775,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2704,
      "step": 36258
    },
    {
      "epoch": 8.631553889186454,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1566,
      "step": 36259
    },
    {
      "epoch": 8.631791941915134,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 36260
    },
    {
      "epoch": 8.632029994643814,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 36261
    },
    {
      "epoch": 8.632268047372493,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9564,
      "step": 36262
    },
    {
      "epoch": 8.632506100101173,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 36263
    },
    {
      "epoch": 8.632744152829853,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1954,
      "step": 36264
    },
    {
      "epoch": 8.63298220555853,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 36265
    },
    {
      "epoch": 8.63322025828721,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 36266
    },
    {
      "epoch": 8.63345831101589,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 36267
    },
    {
      "epoch": 8.63369636374457,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 36268
    },
    {
      "epoch": 8.633934416473249,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 36269
    },
    {
      "epoch": 8.634172469201928,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 36270
    },
    {
      "epoch": 8.634410521930608,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.953,
      "step": 36271
    },
    {
      "epoch": 8.634648574659288,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 36272
    },
    {
      "epoch": 8.634886627387967,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 36273
    },
    {
      "epoch": 8.635124680116645,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9947,
      "step": 36274
    },
    {
      "epoch": 8.635362732845325,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 36275
    },
    {
      "epoch": 8.635600785574004,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8709,
      "step": 36276
    },
    {
      "epoch": 8.635838838302684,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 36277
    },
    {
      "epoch": 8.636076891031363,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3043,
      "step": 36278
    },
    {
      "epoch": 8.636314943760043,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 36279
    },
    {
      "epoch": 8.636552996488723,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.076,
      "step": 36280
    },
    {
      "epoch": 8.636791049217402,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.401,
      "step": 36281
    },
    {
      "epoch": 8.637029101946082,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1702,
      "step": 36282
    },
    {
      "epoch": 8.63726715467476,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0272,
      "step": 36283
    },
    {
      "epoch": 8.63750520740344,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1046,
      "step": 36284
    },
    {
      "epoch": 8.637743260132119,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 36285
    },
    {
      "epoch": 8.637981312860799,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 36286
    },
    {
      "epoch": 8.638219365589478,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 36287
    },
    {
      "epoch": 8.638457418318158,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 36288
    },
    {
      "epoch": 8.638695471046837,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 36289
    },
    {
      "epoch": 8.638933523775517,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 36290
    },
    {
      "epoch": 8.639171576504197,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.6138,
      "step": 36291
    },
    {
      "epoch": 8.639409629232874,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 36292
    },
    {
      "epoch": 8.639647681961554,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1245,
      "step": 36293
    },
    {
      "epoch": 8.639885734690234,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 36294
    },
    {
      "epoch": 8.640123787418913,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 36295
    },
    {
      "epoch": 8.640361840147593,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3812,
      "step": 36296
    },
    {
      "epoch": 8.640599892876272,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 36297
    },
    {
      "epoch": 8.640837945604952,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 36298
    },
    {
      "epoch": 8.641075998333632,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 36299
    },
    {
      "epoch": 8.64131405106231,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2764,
      "step": 36300
    },
    {
      "epoch": 8.64131405106231,
      "eval_runtime": 0.5207,
      "eval_samples_per_second": 1.92,
      "eval_steps_per_second": 1.92,
      "step": 36300
    },
    {
      "epoch": 8.641552103790989,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 36301
    },
    {
      "epoch": 8.641790156519669,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.3754,
      "step": 36302
    },
    {
      "epoch": 8.642028209248348,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 36303
    },
    {
      "epoch": 8.642266261977028,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 36304
    },
    {
      "epoch": 8.642504314705707,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2448,
      "step": 36305
    },
    {
      "epoch": 8.642742367434387,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 36306
    },
    {
      "epoch": 8.642980420163067,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 36307
    },
    {
      "epoch": 8.643218472891746,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 36308
    },
    {
      "epoch": 8.643456525620424,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 36309
    },
    {
      "epoch": 8.643694578349104,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 36310
    },
    {
      "epoch": 8.643932631077783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 36311
    },
    {
      "epoch": 8.644170683806463,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 36312
    },
    {
      "epoch": 8.644408736535143,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 36313
    },
    {
      "epoch": 8.644646789263822,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0321,
      "step": 36314
    },
    {
      "epoch": 8.644884841992502,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1728,
      "step": 36315
    },
    {
      "epoch": 8.645122894721181,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2911,
      "step": 36316
    },
    {
      "epoch": 8.645360947449861,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9449,
      "step": 36317
    },
    {
      "epoch": 8.645599000178539,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 36318
    },
    {
      "epoch": 8.645837052907218,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 36319
    },
    {
      "epoch": 8.646075105635898,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2717,
      "step": 36320
    },
    {
      "epoch": 8.646313158364578,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3387,
      "step": 36321
    },
    {
      "epoch": 8.646551211093257,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 36322
    },
    {
      "epoch": 8.646789263821937,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8885,
      "step": 36323
    },
    {
      "epoch": 8.647027316550616,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 36324
    },
    {
      "epoch": 8.647265369279296,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 36325
    },
    {
      "epoch": 8.647503422007976,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 36326
    },
    {
      "epoch": 8.647741474736653,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 36327
    },
    {
      "epoch": 8.647979527465333,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7104,
      "step": 36328
    },
    {
      "epoch": 8.648217580194013,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8872,
      "step": 36329
    },
    {
      "epoch": 8.648455632922692,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 36330
    },
    {
      "epoch": 8.648693685651372,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 36331
    },
    {
      "epoch": 8.648931738380051,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 36332
    },
    {
      "epoch": 8.649169791108731,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.38,
      "step": 36333
    },
    {
      "epoch": 8.64940784383741,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 36334
    },
    {
      "epoch": 8.649645896566089,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.3424,
      "step": 36335
    },
    {
      "epoch": 8.649883949294768,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 36336
    },
    {
      "epoch": 8.650122002023448,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1313,
      "step": 36337
    },
    {
      "epoch": 8.650360054752127,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 36338
    },
    {
      "epoch": 8.650598107480807,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9202,
      "step": 36339
    },
    {
      "epoch": 8.650836160209487,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 36340
    },
    {
      "epoch": 8.651074212938166,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 36341
    },
    {
      "epoch": 8.651312265666846,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2453,
      "step": 36342
    },
    {
      "epoch": 8.651550318395525,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 36343
    },
    {
      "epoch": 8.651788371124203,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 36344
    },
    {
      "epoch": 8.652026423852883,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 36345
    },
    {
      "epoch": 8.652264476581562,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0836,
      "step": 36346
    },
    {
      "epoch": 8.652502529310242,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 36347
    },
    {
      "epoch": 8.652740582038922,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 36348
    },
    {
      "epoch": 8.652978634767601,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 36349
    },
    {
      "epoch": 8.65321668749628,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1729,
      "step": 36350
    },
    {
      "epoch": 8.65345474022496,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 36351
    },
    {
      "epoch": 8.65369279295364,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 36352
    },
    {
      "epoch": 8.653930845682318,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8291,
      "step": 36353
    },
    {
      "epoch": 8.654168898410997,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8824,
      "step": 36354
    },
    {
      "epoch": 8.654406951139677,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 36355
    },
    {
      "epoch": 8.654645003868357,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 36356
    },
    {
      "epoch": 8.654883056597036,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 36357
    },
    {
      "epoch": 8.655121109325716,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 36358
    },
    {
      "epoch": 8.655359162054395,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 36359
    },
    {
      "epoch": 8.655597214783075,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2739,
      "step": 36360
    },
    {
      "epoch": 8.655835267511755,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8727,
      "step": 36361
    },
    {
      "epoch": 8.656073320240433,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.089,
      "step": 36362
    },
    {
      "epoch": 8.656311372969112,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 36363
    },
    {
      "epoch": 8.656549425697792,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 36364
    },
    {
      "epoch": 8.656787478426471,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0549,
      "step": 36365
    },
    {
      "epoch": 8.657025531155151,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3853,
      "step": 36366
    },
    {
      "epoch": 8.65726358388383,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9519,
      "step": 36367
    },
    {
      "epoch": 8.65750163661251,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 36368
    },
    {
      "epoch": 8.65773968934119,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2333,
      "step": 36369
    },
    {
      "epoch": 8.657977742069868,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 36370
    },
    {
      "epoch": 8.658215794798547,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 36371
    },
    {
      "epoch": 8.658453847527227,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7772,
      "step": 36372
    },
    {
      "epoch": 8.658691900255906,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 36373
    },
    {
      "epoch": 8.658929952984586,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 36374
    },
    {
      "epoch": 8.659168005713266,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2504,
      "step": 36375
    },
    {
      "epoch": 8.659406058441945,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0359,
      "step": 36376
    },
    {
      "epoch": 8.659644111170625,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 36377
    },
    {
      "epoch": 8.659882163899304,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 36378
    },
    {
      "epoch": 8.660120216627984,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 36379
    },
    {
      "epoch": 8.660358269356662,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 36380
    },
    {
      "epoch": 8.660596322085341,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0837,
      "step": 36381
    },
    {
      "epoch": 8.660834374814021,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9752,
      "step": 36382
    },
    {
      "epoch": 8.6610724275427,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2179,
      "step": 36383
    },
    {
      "epoch": 8.66131048027138,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8235,
      "step": 36384
    },
    {
      "epoch": 8.66154853300006,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 36385
    },
    {
      "epoch": 8.66178658572874,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4255,
      "step": 36386
    },
    {
      "epoch": 8.662024638457419,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 36387
    },
    {
      "epoch": 8.662262691186097,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2083,
      "step": 36388
    },
    {
      "epoch": 8.662500743914777,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1281,
      "step": 36389
    },
    {
      "epoch": 8.662738796643456,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 36390
    },
    {
      "epoch": 8.662976849372136,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 36391
    },
    {
      "epoch": 8.663214902100815,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2156,
      "step": 36392
    },
    {
      "epoch": 8.663452954829495,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 36393
    },
    {
      "epoch": 8.663691007558175,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 36394
    },
    {
      "epoch": 8.663929060286854,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9445,
      "step": 36395
    },
    {
      "epoch": 8.664167113015534,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7963,
      "step": 36396
    },
    {
      "epoch": 8.664405165744212,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2702,
      "step": 36397
    },
    {
      "epoch": 8.664643218472891,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.7853,
      "step": 36398
    },
    {
      "epoch": 8.66488127120157,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 36399
    },
    {
      "epoch": 8.66511932393025,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 36400
    },
    {
      "epoch": 8.66511932393025,
      "eval_runtime": 0.5325,
      "eval_samples_per_second": 1.878,
      "eval_steps_per_second": 1.878,
      "step": 36400
    },
    {
      "epoch": 8.66535737665893,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 36401
    },
    {
      "epoch": 8.66559542938761,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 36402
    },
    {
      "epoch": 8.66583348211629,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 36403
    },
    {
      "epoch": 8.666071534844969,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 36404
    },
    {
      "epoch": 8.666309587573647,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 36405
    },
    {
      "epoch": 8.666547640302326,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 36406
    },
    {
      "epoch": 8.666785693031006,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9582,
      "step": 36407
    },
    {
      "epoch": 8.667023745759685,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8972,
      "step": 36408
    },
    {
      "epoch": 8.667261798488365,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9314,
      "step": 36409
    },
    {
      "epoch": 8.667499851217045,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.276,
      "step": 36410
    },
    {
      "epoch": 8.667737903945724,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7803,
      "step": 36411
    },
    {
      "epoch": 8.667975956674404,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 36412
    },
    {
      "epoch": 8.668214009403084,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 36413
    },
    {
      "epoch": 8.668452062131763,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7748,
      "step": 36414
    },
    {
      "epoch": 8.668690114860441,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 36415
    },
    {
      "epoch": 8.66892816758912,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1989,
      "step": 36416
    },
    {
      "epoch": 8.6691662203178,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1056,
      "step": 36417
    },
    {
      "epoch": 8.66940427304648,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 36418
    },
    {
      "epoch": 8.66964232577516,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 36419
    },
    {
      "epoch": 8.669880378503839,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 36420
    },
    {
      "epoch": 8.670118431232519,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 36421
    },
    {
      "epoch": 8.670356483961198,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 36422
    },
    {
      "epoch": 8.670594536689876,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 36423
    },
    {
      "epoch": 8.670832589418556,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 36424
    },
    {
      "epoch": 8.671070642147235,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9462,
      "step": 36425
    },
    {
      "epoch": 8.671308694875915,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0069,
      "step": 36426
    },
    {
      "epoch": 8.671546747604594,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 36427
    },
    {
      "epoch": 8.671784800333274,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3121,
      "step": 36428
    },
    {
      "epoch": 8.672022853061954,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0491,
      "step": 36429
    },
    {
      "epoch": 8.672260905790633,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 36430
    },
    {
      "epoch": 8.672498958519313,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 36431
    },
    {
      "epoch": 8.67273701124799,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.2639,
      "step": 36432
    },
    {
      "epoch": 8.67297506397667,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9594,
      "step": 36433
    },
    {
      "epoch": 8.67321311670535,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9044,
      "step": 36434
    },
    {
      "epoch": 8.67345116943403,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 36435
    },
    {
      "epoch": 8.673689222162709,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 36436
    },
    {
      "epoch": 8.673927274891389,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 36437
    },
    {
      "epoch": 8.674165327620068,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 36438
    },
    {
      "epoch": 8.674403380348748,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7441,
      "step": 36439
    },
    {
      "epoch": 8.674641433077426,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.341,
      "step": 36440
    },
    {
      "epoch": 8.674879485806105,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 36441
    },
    {
      "epoch": 8.675117538534785,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.3586,
      "step": 36442
    },
    {
      "epoch": 8.675355591263465,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 36443
    },
    {
      "epoch": 8.675593643992144,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1377,
      "step": 36444
    },
    {
      "epoch": 8.675831696720824,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 36445
    },
    {
      "epoch": 8.676069749449503,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8204,
      "step": 36446
    },
    {
      "epoch": 8.676307802178183,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9483,
      "step": 36447
    },
    {
      "epoch": 8.676545854906863,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 36448
    },
    {
      "epoch": 8.676783907635542,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2577,
      "step": 36449
    },
    {
      "epoch": 8.67702196036422,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1808,
      "step": 36450
    },
    {
      "epoch": 8.6772600130929,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1108,
      "step": 36451
    },
    {
      "epoch": 8.67749806582158,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 36452
    },
    {
      "epoch": 8.677736118550259,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2025,
      "step": 36453
    },
    {
      "epoch": 8.677974171278938,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1069,
      "step": 36454
    },
    {
      "epoch": 8.678212224007618,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 36455
    },
    {
      "epoch": 8.678450276736298,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 36456
    },
    {
      "epoch": 8.678688329464977,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9626,
      "step": 36457
    },
    {
      "epoch": 8.678926382193655,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 36458
    },
    {
      "epoch": 8.679164434922335,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 36459
    },
    {
      "epoch": 8.679402487651014,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 36460
    },
    {
      "epoch": 8.679640540379694,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 36461
    },
    {
      "epoch": 8.679878593108374,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8774,
      "step": 36462
    },
    {
      "epoch": 8.680116645837053,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 36463
    },
    {
      "epoch": 8.680354698565733,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1735,
      "step": 36464
    },
    {
      "epoch": 8.680592751294412,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.2412,
      "step": 36465
    },
    {
      "epoch": 8.680830804023092,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 36466
    },
    {
      "epoch": 8.68106885675177,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3547,
      "step": 36467
    },
    {
      "epoch": 8.68130690948045,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7472,
      "step": 36468
    },
    {
      "epoch": 8.681544962209129,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2686,
      "step": 36469
    },
    {
      "epoch": 8.681783014937809,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8811,
      "step": 36470
    },
    {
      "epoch": 8.682021067666488,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 36471
    },
    {
      "epoch": 8.682259120395168,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9301,
      "step": 36472
    },
    {
      "epoch": 8.682497173123847,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1999,
      "step": 36473
    },
    {
      "epoch": 8.682735225852527,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 36474
    },
    {
      "epoch": 8.682973278581205,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2254,
      "step": 36475
    },
    {
      "epoch": 8.683211331309884,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1244,
      "step": 36476
    },
    {
      "epoch": 8.683449384038564,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0985,
      "step": 36477
    },
    {
      "epoch": 8.683687436767244,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 36478
    },
    {
      "epoch": 8.683925489495923,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9588,
      "step": 36479
    },
    {
      "epoch": 8.684163542224603,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2162,
      "step": 36480
    },
    {
      "epoch": 8.684401594953282,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 36481
    },
    {
      "epoch": 8.684639647681962,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2336,
      "step": 36482
    },
    {
      "epoch": 8.684877700410642,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 36483
    },
    {
      "epoch": 8.685115753139321,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9391,
      "step": 36484
    },
    {
      "epoch": 8.685353805867999,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 36485
    },
    {
      "epoch": 8.685591858596679,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 36486
    },
    {
      "epoch": 8.685829911325358,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 36487
    },
    {
      "epoch": 8.686067964054038,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 36488
    },
    {
      "epoch": 8.686306016782718,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 36489
    },
    {
      "epoch": 8.686544069511397,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1064,
      "step": 36490
    },
    {
      "epoch": 8.686782122240077,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 36491
    },
    {
      "epoch": 8.687020174968756,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8501,
      "step": 36492
    },
    {
      "epoch": 8.687258227697434,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0017,
      "step": 36493
    },
    {
      "epoch": 8.687496280426114,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 36494
    },
    {
      "epoch": 8.687734333154793,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0391,
      "step": 36495
    },
    {
      "epoch": 8.687972385883473,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 36496
    },
    {
      "epoch": 8.688210438612153,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.423,
      "step": 36497
    },
    {
      "epoch": 8.688448491340832,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 36498
    },
    {
      "epoch": 8.688686544069512,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 36499
    },
    {
      "epoch": 8.688924596798191,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.4006,
      "step": 36500
    },
    {
      "epoch": 8.688924596798191,
      "eval_runtime": 0.5206,
      "eval_samples_per_second": 1.921,
      "eval_steps_per_second": 1.921,
      "step": 36500
    },
    {
      "epoch": 8.689162649526871,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0003,
      "step": 36501
    },
    {
      "epoch": 8.689400702255549,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 36502
    },
    {
      "epoch": 8.689638754984228,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0769,
      "step": 36503
    },
    {
      "epoch": 8.689876807712908,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 36504
    },
    {
      "epoch": 8.690114860441588,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 36505
    },
    {
      "epoch": 8.690352913170267,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.5871,
      "step": 36506
    },
    {
      "epoch": 8.690590965898947,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 36507
    },
    {
      "epoch": 8.690829018627626,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 36508
    },
    {
      "epoch": 8.691067071356306,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 36509
    },
    {
      "epoch": 8.691305124084984,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 36510
    },
    {
      "epoch": 8.691543176813664,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 36511
    },
    {
      "epoch": 8.691781229542343,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 36512
    },
    {
      "epoch": 8.692019282271023,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 36513
    },
    {
      "epoch": 8.692257334999702,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3474,
      "step": 36514
    },
    {
      "epoch": 8.692495387728382,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 36515
    },
    {
      "epoch": 8.692733440457062,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 36516
    },
    {
      "epoch": 8.692971493185741,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 36517
    },
    {
      "epoch": 8.69320954591442,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0106,
      "step": 36518
    },
    {
      "epoch": 8.6934475986431,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2339,
      "step": 36519
    },
    {
      "epoch": 8.693685651371778,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0582,
      "step": 36520
    },
    {
      "epoch": 8.693923704100458,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 36521
    },
    {
      "epoch": 8.694161756829137,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 36522
    },
    {
      "epoch": 8.694399809557817,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 36523
    },
    {
      "epoch": 8.694637862286497,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0452,
      "step": 36524
    },
    {
      "epoch": 8.694875915015176,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 36525
    },
    {
      "epoch": 8.695113967743856,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.6563,
      "step": 36526
    },
    {
      "epoch": 8.695352020472535,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3347,
      "step": 36527
    },
    {
      "epoch": 8.695590073201213,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 36528
    },
    {
      "epoch": 8.695828125929893,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1077,
      "step": 36529
    },
    {
      "epoch": 8.696066178658572,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0368,
      "step": 36530
    },
    {
      "epoch": 8.696304231387252,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 36531
    },
    {
      "epoch": 8.696542284115932,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 36532
    },
    {
      "epoch": 8.696780336844611,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 36533
    },
    {
      "epoch": 8.69701838957329,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2407,
      "step": 36534
    },
    {
      "epoch": 8.69725644230197,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.148,
      "step": 36535
    },
    {
      "epoch": 8.69749449503065,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0004,
      "step": 36536
    },
    {
      "epoch": 8.697732547759328,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 36537
    },
    {
      "epoch": 8.697970600488008,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1569,
      "step": 36538
    },
    {
      "epoch": 8.698208653216687,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 36539
    },
    {
      "epoch": 8.698446705945367,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 36540
    },
    {
      "epoch": 8.698684758674046,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 36541
    },
    {
      "epoch": 8.698922811402726,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 36542
    },
    {
      "epoch": 8.699160864131406,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3525,
      "step": 36543
    },
    {
      "epoch": 8.699398916860085,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2087,
      "step": 36544
    },
    {
      "epoch": 8.699636969588763,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 36545
    },
    {
      "epoch": 8.699875022317443,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0966,
      "step": 36546
    },
    {
      "epoch": 8.700113075046122,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0128,
      "step": 36547
    },
    {
      "epoch": 8.700351127774802,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 36548
    },
    {
      "epoch": 8.700589180503481,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 36549
    },
    {
      "epoch": 8.700827233232161,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3385,
      "step": 36550
    },
    {
      "epoch": 8.70106528596084,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 36551
    },
    {
      "epoch": 8.70130333868952,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 36552
    },
    {
      "epoch": 8.7015413914182,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.6265,
      "step": 36553
    },
    {
      "epoch": 8.70177944414688,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.39,
      "step": 36554
    },
    {
      "epoch": 8.702017496875557,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 36555
    },
    {
      "epoch": 8.702255549604237,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 36556
    },
    {
      "epoch": 8.702493602332916,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0605,
      "step": 36557
    },
    {
      "epoch": 8.702731655061596,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9855,
      "step": 36558
    },
    {
      "epoch": 8.702969707790276,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 36559
    },
    {
      "epoch": 8.703207760518955,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 36560
    },
    {
      "epoch": 8.703445813247635,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 36561
    },
    {
      "epoch": 8.703683865976314,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 36562
    },
    {
      "epoch": 8.703921918704992,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2579,
      "step": 36563
    },
    {
      "epoch": 8.704159971433672,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 36564
    },
    {
      "epoch": 8.704398024162352,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 36565
    },
    {
      "epoch": 8.704636076891031,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 36566
    },
    {
      "epoch": 8.70487412961971,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 36567
    },
    {
      "epoch": 8.70511218234839,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 36568
    },
    {
      "epoch": 8.70535023507707,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8848,
      "step": 36569
    },
    {
      "epoch": 8.70558828780575,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7465,
      "step": 36570
    },
    {
      "epoch": 8.70582634053443,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 36571
    },
    {
      "epoch": 8.706064393263107,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 36572
    },
    {
      "epoch": 8.706302445991787,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 36573
    },
    {
      "epoch": 8.706540498720466,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 36574
    },
    {
      "epoch": 8.706778551449146,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1955,
      "step": 36575
    },
    {
      "epoch": 8.707016604177825,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 36576
    },
    {
      "epoch": 8.707254656906505,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8985,
      "step": 36577
    },
    {
      "epoch": 8.707492709635185,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 36578
    },
    {
      "epoch": 8.707730762363864,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 36579
    },
    {
      "epoch": 8.707968815092542,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 36580
    },
    {
      "epoch": 8.708206867821222,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 36581
    },
    {
      "epoch": 8.708444920549901,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 36582
    },
    {
      "epoch": 8.70868297327858,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 36583
    },
    {
      "epoch": 8.70892102600726,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 36584
    },
    {
      "epoch": 8.70915907873594,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 36585
    },
    {
      "epoch": 8.70939713146462,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 36586
    },
    {
      "epoch": 8.7096351841933,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0247,
      "step": 36587
    },
    {
      "epoch": 8.709873236921979,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 36588
    },
    {
      "epoch": 8.710111289650659,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9863,
      "step": 36589
    },
    {
      "epoch": 8.710349342379336,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 36590
    },
    {
      "epoch": 8.710587395108016,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 36591
    },
    {
      "epoch": 8.710825447836696,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.24,
      "step": 36592
    },
    {
      "epoch": 8.711063500565375,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 36593
    },
    {
      "epoch": 8.711301553294055,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2692,
      "step": 36594
    },
    {
      "epoch": 8.711539606022734,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0882,
      "step": 36595
    },
    {
      "epoch": 8.711777658751414,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 36596
    },
    {
      "epoch": 8.712015711480094,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 36597
    },
    {
      "epoch": 8.712253764208771,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1322,
      "step": 36598
    },
    {
      "epoch": 8.712491816937451,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1144,
      "step": 36599
    },
    {
      "epoch": 8.71272986966613,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9598,
      "step": 36600
    },
    {
      "epoch": 8.71272986966613,
      "eval_runtime": 0.5239,
      "eval_samples_per_second": 1.909,
      "eval_steps_per_second": 1.909,
      "step": 36600
    },
    {
      "epoch": 8.71296792239481,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 36601
    },
    {
      "epoch": 8.71320597512349,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 36602
    },
    {
      "epoch": 8.71344402785217,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 36603
    },
    {
      "epoch": 8.713682080580849,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1958,
      "step": 36604
    },
    {
      "epoch": 8.713920133309529,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8326,
      "step": 36605
    },
    {
      "epoch": 8.714158186038208,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 36606
    },
    {
      "epoch": 8.714396238766886,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 36607
    },
    {
      "epoch": 8.714634291495566,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 36608
    },
    {
      "epoch": 8.714872344224245,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1618,
      "step": 36609
    },
    {
      "epoch": 8.715110396952925,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 36610
    },
    {
      "epoch": 8.715348449681604,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 36611
    },
    {
      "epoch": 8.715586502410284,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2085,
      "step": 36612
    },
    {
      "epoch": 8.715824555138964,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9492,
      "step": 36613
    },
    {
      "epoch": 8.716062607867643,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 36614
    },
    {
      "epoch": 8.716300660596323,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8354,
      "step": 36615
    },
    {
      "epoch": 8.716538713325,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2236,
      "step": 36616
    },
    {
      "epoch": 8.71677676605368,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 36617
    },
    {
      "epoch": 8.71701481878236,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0055,
      "step": 36618
    },
    {
      "epoch": 8.71725287151104,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 36619
    },
    {
      "epoch": 8.71749092423972,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2033,
      "step": 36620
    },
    {
      "epoch": 8.717728976968399,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 36621
    },
    {
      "epoch": 8.717967029697078,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 36622
    },
    {
      "epoch": 8.718205082425758,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.6969,
      "step": 36623
    },
    {
      "epoch": 8.718443135154438,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 36624
    },
    {
      "epoch": 8.718681187883115,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2092,
      "step": 36625
    },
    {
      "epoch": 8.718919240611795,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 36626
    },
    {
      "epoch": 8.719157293340475,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 36627
    },
    {
      "epoch": 8.719395346069154,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 36628
    },
    {
      "epoch": 8.719633398797834,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.246,
      "step": 36629
    },
    {
      "epoch": 8.719871451526513,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8184,
      "step": 36630
    },
    {
      "epoch": 8.720109504255193,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.7187,
      "step": 36631
    },
    {
      "epoch": 8.720347556983873,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.4935,
      "step": 36632
    },
    {
      "epoch": 8.72058560971255,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2703,
      "step": 36633
    },
    {
      "epoch": 8.72082366244123,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1877,
      "step": 36634
    },
    {
      "epoch": 8.72106171516991,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1423,
      "step": 36635
    },
    {
      "epoch": 8.72129976789859,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1199,
      "step": 36636
    },
    {
      "epoch": 8.721537820627269,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 36637
    },
    {
      "epoch": 8.721775873355949,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 36638
    },
    {
      "epoch": 8.722013926084628,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4215,
      "step": 36639
    },
    {
      "epoch": 8.722251978813308,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1893,
      "step": 36640
    },
    {
      "epoch": 8.722490031541987,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 36641
    },
    {
      "epoch": 8.722728084270665,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2666,
      "step": 36642
    },
    {
      "epoch": 8.722966136999345,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 36643
    },
    {
      "epoch": 8.723204189728024,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 36644
    },
    {
      "epoch": 8.723442242456704,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.1392,
      "step": 36645
    },
    {
      "epoch": 8.723680295185384,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1997,
      "step": 36646
    },
    {
      "epoch": 8.723918347914063,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9451,
      "step": 36647
    },
    {
      "epoch": 8.724156400642743,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0509,
      "step": 36648
    },
    {
      "epoch": 8.724394453371422,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8684,
      "step": 36649
    },
    {
      "epoch": 8.724632506100102,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 36650
    },
    {
      "epoch": 8.72487055882878,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 36651
    },
    {
      "epoch": 8.72510861155746,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 36652
    },
    {
      "epoch": 8.725346664286139,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 36653
    },
    {
      "epoch": 8.725584717014819,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8206,
      "step": 36654
    },
    {
      "epoch": 8.725822769743498,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 36655
    },
    {
      "epoch": 8.726060822472178,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 36656
    },
    {
      "epoch": 8.726298875200857,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9589,
      "step": 36657
    },
    {
      "epoch": 8.726536927929537,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8962,
      "step": 36658
    },
    {
      "epoch": 8.726774980658217,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4077,
      "step": 36659
    },
    {
      "epoch": 8.727013033386894,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3731,
      "step": 36660
    },
    {
      "epoch": 8.727251086115574,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1979,
      "step": 36661
    },
    {
      "epoch": 8.727489138844254,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8564,
      "step": 36662
    },
    {
      "epoch": 8.727727191572933,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 36663
    },
    {
      "epoch": 8.727965244301613,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 36664
    },
    {
      "epoch": 8.728203297030293,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 36665
    },
    {
      "epoch": 8.728441349758972,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 36666
    },
    {
      "epoch": 8.728679402487652,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1277,
      "step": 36667
    },
    {
      "epoch": 8.72891745521633,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0482,
      "step": 36668
    },
    {
      "epoch": 8.72915550794501,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 36669
    },
    {
      "epoch": 8.729393560673689,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 36670
    },
    {
      "epoch": 8.729631613402368,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 36671
    },
    {
      "epoch": 8.729869666131048,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2564,
      "step": 36672
    },
    {
      "epoch": 8.730107718859728,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 36673
    },
    {
      "epoch": 8.730345771588407,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.43,
      "step": 36674
    },
    {
      "epoch": 8.730583824317087,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1697,
      "step": 36675
    },
    {
      "epoch": 8.730821877045766,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8127,
      "step": 36676
    },
    {
      "epoch": 8.731059929774444,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 36677
    },
    {
      "epoch": 8.731297982503124,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 36678
    },
    {
      "epoch": 8.731536035231803,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2535,
      "step": 36679
    },
    {
      "epoch": 8.731774087960483,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 36680
    },
    {
      "epoch": 8.732012140689163,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 36681
    },
    {
      "epoch": 8.732250193417842,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 36682
    },
    {
      "epoch": 8.732488246146522,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 36683
    },
    {
      "epoch": 8.732726298875201,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9671,
      "step": 36684
    },
    {
      "epoch": 8.732964351603881,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2243,
      "step": 36685
    },
    {
      "epoch": 8.733202404332559,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 36686
    },
    {
      "epoch": 8.733440457061239,
      "grad_norm": 2.78125,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 36687
    },
    {
      "epoch": 8.733678509789918,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 36688
    },
    {
      "epoch": 8.733916562518598,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 36689
    },
    {
      "epoch": 8.734154615247277,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 36690
    },
    {
      "epoch": 8.734392667975957,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 36691
    },
    {
      "epoch": 8.734630720704637,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 36692
    },
    {
      "epoch": 8.734868773433316,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 36693
    },
    {
      "epoch": 8.735106826161996,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8595,
      "step": 36694
    },
    {
      "epoch": 8.735344878890674,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9032,
      "step": 36695
    },
    {
      "epoch": 8.735582931619353,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1406,
      "step": 36696
    },
    {
      "epoch": 8.735820984348033,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0993,
      "step": 36697
    },
    {
      "epoch": 8.736059037076712,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0229,
      "step": 36698
    },
    {
      "epoch": 8.736297089805392,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3886,
      "step": 36699
    },
    {
      "epoch": 8.736535142534072,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 36700
    },
    {
      "epoch": 8.736535142534072,
      "eval_runtime": 0.5175,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 36700
    },
    {
      "epoch": 8.736773195262751,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8859,
      "step": 36701
    },
    {
      "epoch": 8.73701124799143,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9948,
      "step": 36702
    },
    {
      "epoch": 8.737249300720109,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 36703
    },
    {
      "epoch": 8.737487353448788,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.159,
      "step": 36704
    },
    {
      "epoch": 8.737725406177468,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 36705
    },
    {
      "epoch": 8.737963458906147,
      "grad_norm": 5.75,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 36706
    },
    {
      "epoch": 8.738201511634827,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 36707
    },
    {
      "epoch": 8.738439564363507,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 36708
    },
    {
      "epoch": 8.738677617092186,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 36709
    },
    {
      "epoch": 8.738915669820866,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 36710
    },
    {
      "epoch": 8.739153722549545,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2091,
      "step": 36711
    },
    {
      "epoch": 8.739391775278223,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 36712
    },
    {
      "epoch": 8.739629828006903,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 36713
    },
    {
      "epoch": 8.739867880735583,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 36714
    },
    {
      "epoch": 8.740105933464262,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1567,
      "step": 36715
    },
    {
      "epoch": 8.740343986192942,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2702,
      "step": 36716
    },
    {
      "epoch": 8.740582038921621,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 36717
    },
    {
      "epoch": 8.740820091650301,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 36718
    },
    {
      "epoch": 8.74105814437898,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 36719
    },
    {
      "epoch": 8.74129619710766,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8246,
      "step": 36720
    },
    {
      "epoch": 8.741534249836338,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 36721
    },
    {
      "epoch": 8.741772302565018,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 36722
    },
    {
      "epoch": 8.742010355293697,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 36723
    },
    {
      "epoch": 8.742248408022377,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0414,
      "step": 36724
    },
    {
      "epoch": 8.742486460751056,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.445,
      "step": 36725
    },
    {
      "epoch": 8.742724513479736,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9295,
      "step": 36726
    },
    {
      "epoch": 8.742962566208416,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 36727
    },
    {
      "epoch": 8.743200618937095,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 36728
    },
    {
      "epoch": 8.743438671665775,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 36729
    },
    {
      "epoch": 8.743676724394453,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2701,
      "step": 36730
    },
    {
      "epoch": 8.743914777123132,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 36731
    },
    {
      "epoch": 8.744152829851812,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 36732
    },
    {
      "epoch": 8.744390882580491,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 36733
    },
    {
      "epoch": 8.744628935309171,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0033,
      "step": 36734
    },
    {
      "epoch": 8.74486698803785,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9144,
      "step": 36735
    },
    {
      "epoch": 8.74510504076653,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8977,
      "step": 36736
    },
    {
      "epoch": 8.74534309349521,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8724,
      "step": 36737
    },
    {
      "epoch": 8.745581146223888,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 36738
    },
    {
      "epoch": 8.745819198952567,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0898,
      "step": 36739
    },
    {
      "epoch": 8.746057251681247,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3844,
      "step": 36740
    },
    {
      "epoch": 8.746295304409927,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 36741
    },
    {
      "epoch": 8.746533357138606,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 36742
    },
    {
      "epoch": 8.746771409867286,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 36743
    },
    {
      "epoch": 8.747009462595965,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1987,
      "step": 36744
    },
    {
      "epoch": 8.747247515324645,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 36745
    },
    {
      "epoch": 8.747485568053325,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 36746
    },
    {
      "epoch": 8.747723620782002,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 36747
    },
    {
      "epoch": 8.747961673510682,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.5386,
      "step": 36748
    },
    {
      "epoch": 8.748199726239362,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2086,
      "step": 36749
    },
    {
      "epoch": 8.748437778968041,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 36750
    },
    {
      "epoch": 8.74867583169672,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8654,
      "step": 36751
    },
    {
      "epoch": 8.7489138844254,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 36752
    },
    {
      "epoch": 8.74915193715408,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7809,
      "step": 36753
    },
    {
      "epoch": 8.74938998988276,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 36754
    },
    {
      "epoch": 8.74962804261144,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 36755
    },
    {
      "epoch": 8.749866095340117,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1897,
      "step": 36756
    },
    {
      "epoch": 8.750104148068797,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 36757
    },
    {
      "epoch": 8.750342200797476,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 36758
    },
    {
      "epoch": 8.750580253526156,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9341,
      "step": 36759
    },
    {
      "epoch": 8.750818306254835,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 36760
    },
    {
      "epoch": 8.751056358983515,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 36761
    },
    {
      "epoch": 8.751294411712195,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 36762
    },
    {
      "epoch": 8.751532464440874,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 36763
    },
    {
      "epoch": 8.751770517169554,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9543,
      "step": 36764
    },
    {
      "epoch": 8.752008569898232,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 36765
    },
    {
      "epoch": 8.752246622626911,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0859,
      "step": 36766
    },
    {
      "epoch": 8.752484675355591,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 36767
    },
    {
      "epoch": 8.75272272808427,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 36768
    },
    {
      "epoch": 8.75296078081295,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 36769
    },
    {
      "epoch": 8.75319883354163,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 36770
    },
    {
      "epoch": 8.75343688627031,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.8521,
      "step": 36771
    },
    {
      "epoch": 8.753674938998989,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9087,
      "step": 36772
    },
    {
      "epoch": 8.753912991727667,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 36773
    },
    {
      "epoch": 8.754151044456346,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 36774
    },
    {
      "epoch": 8.754389097185026,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0725,
      "step": 36775
    },
    {
      "epoch": 8.754627149913706,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 36776
    },
    {
      "epoch": 8.754865202642385,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1364,
      "step": 36777
    },
    {
      "epoch": 8.755103255371065,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9928,
      "step": 36778
    },
    {
      "epoch": 8.755341308099744,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 36779
    },
    {
      "epoch": 8.755579360828424,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 36780
    },
    {
      "epoch": 8.755817413557104,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8082,
      "step": 36781
    },
    {
      "epoch": 8.756055466285783,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.4987,
      "step": 36782
    },
    {
      "epoch": 8.756293519014461,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 36783
    },
    {
      "epoch": 8.75653157174314,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 36784
    },
    {
      "epoch": 8.75676962447182,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0896,
      "step": 36785
    },
    {
      "epoch": 8.7570076772005,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 36786
    },
    {
      "epoch": 8.75724572992918,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 36787
    },
    {
      "epoch": 8.757483782657859,
      "grad_norm": 2.84375,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 36788
    },
    {
      "epoch": 8.757721835386539,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2297,
      "step": 36789
    },
    {
      "epoch": 8.757959888115218,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 36790
    },
    {
      "epoch": 8.758197940843896,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1439,
      "step": 36791
    },
    {
      "epoch": 8.758435993572576,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 36792
    },
    {
      "epoch": 8.758674046301255,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.167,
      "step": 36793
    },
    {
      "epoch": 8.758912099029935,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 36794
    },
    {
      "epoch": 8.759150151758615,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0876,
      "step": 36795
    },
    {
      "epoch": 8.759388204487294,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 36796
    },
    {
      "epoch": 8.759626257215974,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 36797
    },
    {
      "epoch": 8.759864309944653,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7861,
      "step": 36798
    },
    {
      "epoch": 8.760102362673333,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8281,
      "step": 36799
    },
    {
      "epoch": 8.76034041540201,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1494,
      "step": 36800
    },
    {
      "epoch": 8.76034041540201,
      "eval_runtime": 0.5139,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 36800
    },
    {
      "epoch": 8.76057846813069,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 36801
    },
    {
      "epoch": 8.76081652085937,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 36802
    },
    {
      "epoch": 8.76105457358805,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9915,
      "step": 36803
    },
    {
      "epoch": 8.76129262631673,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9711,
      "step": 36804
    },
    {
      "epoch": 8.761530679045409,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 36805
    },
    {
      "epoch": 8.761768731774088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 36806
    },
    {
      "epoch": 8.762006784502768,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 36807
    },
    {
      "epoch": 8.762244837231446,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 36808
    },
    {
      "epoch": 8.762482889960125,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1459,
      "step": 36809
    },
    {
      "epoch": 8.762720942688805,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8227,
      "step": 36810
    },
    {
      "epoch": 8.762958995417485,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 36811
    },
    {
      "epoch": 8.763197048146164,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 36812
    },
    {
      "epoch": 8.763435100874844,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 36813
    },
    {
      "epoch": 8.763673153603524,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8826,
      "step": 36814
    },
    {
      "epoch": 8.763911206332203,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 36815
    },
    {
      "epoch": 8.764149259060883,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.814,
      "step": 36816
    },
    {
      "epoch": 8.764387311789562,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 36817
    },
    {
      "epoch": 8.76462536451824,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 36818
    },
    {
      "epoch": 8.76486341724692,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1228,
      "step": 36819
    },
    {
      "epoch": 8.7651014699756,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0915,
      "step": 36820
    },
    {
      "epoch": 8.765339522704279,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 36821
    },
    {
      "epoch": 8.765577575432959,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 36822
    },
    {
      "epoch": 8.765815628161638,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 36823
    },
    {
      "epoch": 8.766053680890318,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 36824
    },
    {
      "epoch": 8.766291733618997,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1898,
      "step": 36825
    },
    {
      "epoch": 8.766529786347675,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 36826
    },
    {
      "epoch": 8.766767839076355,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 36827
    },
    {
      "epoch": 8.767005891805034,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 36828
    },
    {
      "epoch": 8.767243944533714,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3096,
      "step": 36829
    },
    {
      "epoch": 8.767481997262394,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2344,
      "step": 36830
    },
    {
      "epoch": 8.767720049991073,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.813,
      "step": 36831
    },
    {
      "epoch": 8.767958102719753,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 36832
    },
    {
      "epoch": 8.768196155448432,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 36833
    },
    {
      "epoch": 8.768434208177112,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 36834
    },
    {
      "epoch": 8.76867226090579,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3688,
      "step": 36835
    },
    {
      "epoch": 8.76891031363447,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 36836
    },
    {
      "epoch": 8.769148366363149,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4031,
      "step": 36837
    },
    {
      "epoch": 8.769386419091829,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 36838
    },
    {
      "epoch": 8.769624471820508,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1302,
      "step": 36839
    },
    {
      "epoch": 8.769862524549188,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8894,
      "step": 36840
    },
    {
      "epoch": 8.770100577277868,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.2596,
      "step": 36841
    },
    {
      "epoch": 8.770338630006547,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 36842
    },
    {
      "epoch": 8.770576682735225,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 36843
    },
    {
      "epoch": 8.770814735463905,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.923,
      "step": 36844
    },
    {
      "epoch": 8.771052788192584,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 36845
    },
    {
      "epoch": 8.771290840921264,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.99,
      "step": 36846
    },
    {
      "epoch": 8.771528893649943,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 36847
    },
    {
      "epoch": 8.771766946378623,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0545,
      "step": 36848
    },
    {
      "epoch": 8.772004999107303,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.081,
      "step": 36849
    },
    {
      "epoch": 8.772243051835982,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.2103,
      "step": 36850
    },
    {
      "epoch": 8.772481104564662,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 36851
    },
    {
      "epoch": 8.772719157293341,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 36852
    },
    {
      "epoch": 8.77295721002202,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1972,
      "step": 36853
    },
    {
      "epoch": 8.773195262750699,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 36854
    },
    {
      "epoch": 8.773433315479378,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 36855
    },
    {
      "epoch": 8.773671368208058,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 36856
    },
    {
      "epoch": 8.773909420936738,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 36857
    },
    {
      "epoch": 8.774147473665417,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 36858
    },
    {
      "epoch": 8.774385526394097,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 36859
    },
    {
      "epoch": 8.774623579122776,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1354,
      "step": 36860
    },
    {
      "epoch": 8.774861631851454,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 36861
    },
    {
      "epoch": 8.775099684580134,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1452,
      "step": 36862
    },
    {
      "epoch": 8.775337737308814,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2971,
      "step": 36863
    },
    {
      "epoch": 8.775575790037493,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 36864
    },
    {
      "epoch": 8.775813842766173,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1734,
      "step": 36865
    },
    {
      "epoch": 8.776051895494852,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 36866
    },
    {
      "epoch": 8.776289948223532,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2239,
      "step": 36867
    },
    {
      "epoch": 8.776528000952212,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.4261,
      "step": 36868
    },
    {
      "epoch": 8.776766053680891,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 36869
    },
    {
      "epoch": 8.777004106409569,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9289,
      "step": 36870
    },
    {
      "epoch": 8.777242159138249,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 36871
    },
    {
      "epoch": 8.777480211866928,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 36872
    },
    {
      "epoch": 8.777718264595608,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1973,
      "step": 36873
    },
    {
      "epoch": 8.777956317324287,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1226,
      "step": 36874
    },
    {
      "epoch": 8.778194370052967,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 36875
    },
    {
      "epoch": 8.778432422781647,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0416,
      "step": 36876
    },
    {
      "epoch": 8.778670475510326,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 36877
    },
    {
      "epoch": 8.778908528239004,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1879,
      "step": 36878
    },
    {
      "epoch": 8.779146580967684,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 36879
    },
    {
      "epoch": 8.779384633696363,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 36880
    },
    {
      "epoch": 8.779622686425043,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 36881
    },
    {
      "epoch": 8.779860739153722,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0222,
      "step": 36882
    },
    {
      "epoch": 8.780098791882402,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9313,
      "step": 36883
    },
    {
      "epoch": 8.780336844611082,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1906,
      "step": 36884
    },
    {
      "epoch": 8.780574897339761,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 36885
    },
    {
      "epoch": 8.78081295006844,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 36886
    },
    {
      "epoch": 8.78105100279712,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 36887
    },
    {
      "epoch": 8.781289055525798,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 36888
    },
    {
      "epoch": 8.781527108254478,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2731,
      "step": 36889
    },
    {
      "epoch": 8.781765160983158,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 36890
    },
    {
      "epoch": 8.782003213711837,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 36891
    },
    {
      "epoch": 8.782241266440517,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9189,
      "step": 36892
    },
    {
      "epoch": 8.782479319169196,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6814,
      "step": 36893
    },
    {
      "epoch": 8.782717371897876,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1892,
      "step": 36894
    },
    {
      "epoch": 8.782955424626556,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 36895
    },
    {
      "epoch": 8.783193477355233,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1545,
      "step": 36896
    },
    {
      "epoch": 8.783431530083913,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 36897
    },
    {
      "epoch": 8.783669582812593,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 36898
    },
    {
      "epoch": 8.783907635541272,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 36899
    },
    {
      "epoch": 8.784145688269952,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1334,
      "step": 36900
    },
    {
      "epoch": 8.784145688269952,
      "eval_runtime": 0.5314,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 36900
    },
    {
      "epoch": 8.784383740998631,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 36901
    },
    {
      "epoch": 8.784621793727311,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8996,
      "step": 36902
    },
    {
      "epoch": 8.78485984645599,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9557,
      "step": 36903
    },
    {
      "epoch": 8.78509789918467,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 36904
    },
    {
      "epoch": 8.785335951913348,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0402,
      "step": 36905
    },
    {
      "epoch": 8.785574004642028,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 36906
    },
    {
      "epoch": 8.785812057370707,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 36907
    },
    {
      "epoch": 8.786050110099387,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 36908
    },
    {
      "epoch": 8.786288162828066,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 36909
    },
    {
      "epoch": 8.786526215556746,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 36910
    },
    {
      "epoch": 8.786764268285426,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9662,
      "step": 36911
    },
    {
      "epoch": 8.787002321014105,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2325,
      "step": 36912
    },
    {
      "epoch": 8.787240373742783,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9913,
      "step": 36913
    },
    {
      "epoch": 8.787478426471463,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 36914
    },
    {
      "epoch": 8.787716479200142,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8243,
      "step": 36915
    },
    {
      "epoch": 8.787954531928822,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 36916
    },
    {
      "epoch": 8.788192584657502,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3651,
      "step": 36917
    },
    {
      "epoch": 8.788430637386181,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 36918
    },
    {
      "epoch": 8.78866869011486,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7699,
      "step": 36919
    },
    {
      "epoch": 8.78890674284354,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 36920
    },
    {
      "epoch": 8.78914479557222,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 36921
    },
    {
      "epoch": 8.7893828483009,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 36922
    },
    {
      "epoch": 8.789620901029577,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9566,
      "step": 36923
    },
    {
      "epoch": 8.789858953758257,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8609,
      "step": 36924
    },
    {
      "epoch": 8.790097006486937,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 36925
    },
    {
      "epoch": 8.790335059215616,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 36926
    },
    {
      "epoch": 8.790573111944296,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 36927
    },
    {
      "epoch": 8.790811164672975,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2809,
      "step": 36928
    },
    {
      "epoch": 8.791049217401655,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 36929
    },
    {
      "epoch": 8.791287270130335,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 36930
    },
    {
      "epoch": 8.791525322859012,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.3833,
      "step": 36931
    },
    {
      "epoch": 8.791763375587692,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 36932
    },
    {
      "epoch": 8.792001428316372,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8361,
      "step": 36933
    },
    {
      "epoch": 8.792239481045051,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1742,
      "step": 36934
    },
    {
      "epoch": 8.79247753377373,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0219,
      "step": 36935
    },
    {
      "epoch": 8.79271558650241,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 36936
    },
    {
      "epoch": 8.79295363923109,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.7514,
      "step": 36937
    },
    {
      "epoch": 8.79319169195977,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 36938
    },
    {
      "epoch": 8.79342974468845,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 36939
    },
    {
      "epoch": 8.793667797417127,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1095,
      "step": 36940
    },
    {
      "epoch": 8.793905850145807,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9075,
      "step": 36941
    },
    {
      "epoch": 8.794143902874486,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 36942
    },
    {
      "epoch": 8.794381955603166,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 36943
    },
    {
      "epoch": 8.794620008331846,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0633,
      "step": 36944
    },
    {
      "epoch": 8.794858061060525,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 36945
    },
    {
      "epoch": 8.795096113789205,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 36946
    },
    {
      "epoch": 8.795334166517884,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0236,
      "step": 36947
    },
    {
      "epoch": 8.795572219246562,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 36948
    },
    {
      "epoch": 8.795810271975242,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 36949
    },
    {
      "epoch": 8.796048324703921,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.3786,
      "step": 36950
    },
    {
      "epoch": 8.796286377432601,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 36951
    },
    {
      "epoch": 8.79652443016128,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7604,
      "step": 36952
    },
    {
      "epoch": 8.79676248288996,
      "grad_norm": 2.9375,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 36953
    },
    {
      "epoch": 8.79700053561864,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 36954
    },
    {
      "epoch": 8.79723858834732,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 36955
    },
    {
      "epoch": 8.797476641075999,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 36956
    },
    {
      "epoch": 8.797714693804679,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 36957
    },
    {
      "epoch": 8.797952746533356,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0973,
      "step": 36958
    },
    {
      "epoch": 8.798190799262036,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1942,
      "step": 36959
    },
    {
      "epoch": 8.798428851990716,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 36960
    },
    {
      "epoch": 8.798666904719395,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1657,
      "step": 36961
    },
    {
      "epoch": 8.798904957448075,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 36962
    },
    {
      "epoch": 8.799143010176754,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2387,
      "step": 36963
    },
    {
      "epoch": 8.799381062905434,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0431,
      "step": 36964
    },
    {
      "epoch": 8.799619115634114,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9643,
      "step": 36965
    },
    {
      "epoch": 8.799857168362792,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 36966
    },
    {
      "epoch": 8.800095221091471,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3561,
      "step": 36967
    },
    {
      "epoch": 8.80033327382015,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9292,
      "step": 36968
    },
    {
      "epoch": 8.80057132654883,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7955,
      "step": 36969
    },
    {
      "epoch": 8.80080937927751,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 36970
    },
    {
      "epoch": 8.80104743200619,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 36971
    },
    {
      "epoch": 8.80128548473487,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.8426,
      "step": 36972
    },
    {
      "epoch": 8.801523537463549,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0756,
      "step": 36973
    },
    {
      "epoch": 8.801761590192228,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 36974
    },
    {
      "epoch": 8.801999642920906,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0433,
      "step": 36975
    },
    {
      "epoch": 8.802237695649586,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0975,
      "step": 36976
    },
    {
      "epoch": 8.802475748378265,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0211,
      "step": 36977
    },
    {
      "epoch": 8.802713801106945,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 36978
    },
    {
      "epoch": 8.802951853835625,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2943,
      "step": 36979
    },
    {
      "epoch": 8.803189906564304,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 36980
    },
    {
      "epoch": 8.803427959292984,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "step": 36981
    },
    {
      "epoch": 8.803666012021663,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8611,
      "step": 36982
    },
    {
      "epoch": 8.803904064750341,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 36983
    },
    {
      "epoch": 8.80414211747902,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9086,
      "step": 36984
    },
    {
      "epoch": 8.8043801702077,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 36985
    },
    {
      "epoch": 8.80461822293638,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0979,
      "step": 36986
    },
    {
      "epoch": 8.80485627566506,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7717,
      "step": 36987
    },
    {
      "epoch": 8.80509432839374,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9731,
      "step": 36988
    },
    {
      "epoch": 8.805332381122419,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 36989
    },
    {
      "epoch": 8.805570433851098,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.6996,
      "step": 36990
    },
    {
      "epoch": 8.805808486579778,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9937,
      "step": 36991
    },
    {
      "epoch": 8.806046539308458,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 36992
    },
    {
      "epoch": 8.806284592037136,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.5864,
      "step": 36993
    },
    {
      "epoch": 8.806522644765815,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 36994
    },
    {
      "epoch": 8.806760697494495,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 36995
    },
    {
      "epoch": 8.806998750223174,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 36996
    },
    {
      "epoch": 8.807236802951854,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.998,
      "step": 36997
    },
    {
      "epoch": 8.807474855680534,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0739,
      "step": 36998
    },
    {
      "epoch": 8.807712908409213,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 36999
    },
    {
      "epoch": 8.807950961137893,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.11,
      "step": 37000
    },
    {
      "epoch": 8.807950961137893,
      "eval_runtime": 0.5304,
      "eval_samples_per_second": 1.885,
      "eval_steps_per_second": 1.885,
      "step": 37000
    },
    {
      "epoch": 8.80818901386657,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.8329,
      "step": 37001
    },
    {
      "epoch": 8.80842706659525,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.4349,
      "step": 37002
    },
    {
      "epoch": 8.80866511932393,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 37003
    },
    {
      "epoch": 8.80890317205261,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9045,
      "step": 37004
    },
    {
      "epoch": 8.809141224781289,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.2098,
      "step": 37005
    },
    {
      "epoch": 8.809379277509969,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 37006
    },
    {
      "epoch": 8.809617330238648,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 37007
    },
    {
      "epoch": 8.809855382967328,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 37008
    },
    {
      "epoch": 8.810093435696007,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 37009
    },
    {
      "epoch": 8.810331488424685,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 37010
    },
    {
      "epoch": 8.810569541153365,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 37011
    },
    {
      "epoch": 8.810807593882044,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.929,
      "step": 37012
    },
    {
      "epoch": 8.811045646610724,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1899,
      "step": 37013
    },
    {
      "epoch": 8.811283699339404,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1181,
      "step": 37014
    },
    {
      "epoch": 8.811521752068083,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 37015
    },
    {
      "epoch": 8.811759804796763,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 37016
    },
    {
      "epoch": 8.811997857525443,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 37017
    },
    {
      "epoch": 8.812235910254122,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2062,
      "step": 37018
    },
    {
      "epoch": 8.8124739629828,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 37019
    },
    {
      "epoch": 8.81271201571148,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 37020
    },
    {
      "epoch": 8.81295006844016,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 37021
    },
    {
      "epoch": 8.813188121168839,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0864,
      "step": 37022
    },
    {
      "epoch": 8.813426173897518,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.973,
      "step": 37023
    },
    {
      "epoch": 8.813664226626198,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1089,
      "step": 37024
    },
    {
      "epoch": 8.813902279354878,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 37025
    },
    {
      "epoch": 8.814140332083557,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 37026
    },
    {
      "epoch": 8.814378384812237,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7593,
      "step": 37027
    },
    {
      "epoch": 8.814616437540915,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 37028
    },
    {
      "epoch": 8.814854490269594,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1711,
      "step": 37029
    },
    {
      "epoch": 8.815092542998274,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 37030
    },
    {
      "epoch": 8.815330595726953,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8925,
      "step": 37031
    },
    {
      "epoch": 8.815568648455633,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 37032
    },
    {
      "epoch": 8.815806701184313,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1924,
      "step": 37033
    },
    {
      "epoch": 8.816044753912992,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 37034
    },
    {
      "epoch": 8.816282806641672,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9478,
      "step": 37035
    },
    {
      "epoch": 8.81652085937035,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 37036
    },
    {
      "epoch": 8.81675891209903,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 37037
    },
    {
      "epoch": 8.816996964827709,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 37038
    },
    {
      "epoch": 8.817235017556388,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 37039
    },
    {
      "epoch": 8.817473070285068,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 37040
    },
    {
      "epoch": 8.817711123013748,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 37041
    },
    {
      "epoch": 8.817949175742427,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 37042
    },
    {
      "epoch": 8.818187228471107,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 37043
    },
    {
      "epoch": 8.818425281199787,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 37044
    },
    {
      "epoch": 8.818663333928464,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 37045
    },
    {
      "epoch": 8.818901386657144,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 37046
    },
    {
      "epoch": 8.819139439385824,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9899,
      "step": 37047
    },
    {
      "epoch": 8.819377492114503,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3661,
      "step": 37048
    },
    {
      "epoch": 8.819615544843183,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1202,
      "step": 37049
    },
    {
      "epoch": 8.819853597571862,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 37050
    },
    {
      "epoch": 8.820091650300542,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0826,
      "step": 37051
    },
    {
      "epoch": 8.820329703029222,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 37052
    },
    {
      "epoch": 8.820567755757901,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0175,
      "step": 37053
    },
    {
      "epoch": 8.820805808486579,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4022,
      "step": 37054
    },
    {
      "epoch": 8.821043861215259,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 37055
    },
    {
      "epoch": 8.821281913943938,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9245,
      "step": 37056
    },
    {
      "epoch": 8.821519966672618,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1588,
      "step": 37057
    },
    {
      "epoch": 8.821758019401297,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2924,
      "step": 37058
    },
    {
      "epoch": 8.821996072129977,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1292,
      "step": 37059
    },
    {
      "epoch": 8.822234124858657,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 37060
    },
    {
      "epoch": 8.822472177587336,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 37061
    },
    {
      "epoch": 8.822710230316016,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0817,
      "step": 37062
    },
    {
      "epoch": 8.822948283044694,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 37063
    },
    {
      "epoch": 8.823186335773373,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2514,
      "step": 37064
    },
    {
      "epoch": 8.823424388502053,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 37065
    },
    {
      "epoch": 8.823662441230733,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 37066
    },
    {
      "epoch": 8.823900493959412,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 37067
    },
    {
      "epoch": 8.824138546688092,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.3146,
      "step": 37068
    },
    {
      "epoch": 8.824376599416771,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 37069
    },
    {
      "epoch": 8.824614652145451,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0791,
      "step": 37070
    },
    {
      "epoch": 8.824852704874129,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 37071
    },
    {
      "epoch": 8.825090757602808,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 37072
    },
    {
      "epoch": 8.825328810331488,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8995,
      "step": 37073
    },
    {
      "epoch": 8.825566863060168,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 37074
    },
    {
      "epoch": 8.825804915788847,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 37075
    },
    {
      "epoch": 8.826042968517527,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9573,
      "step": 37076
    },
    {
      "epoch": 8.826281021246206,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 37077
    },
    {
      "epoch": 8.826519073974886,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 37078
    },
    {
      "epoch": 8.826757126703566,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 37079
    },
    {
      "epoch": 8.826995179432243,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 37080
    },
    {
      "epoch": 8.827233232160923,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 37081
    },
    {
      "epoch": 8.827471284889603,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 37082
    },
    {
      "epoch": 8.827709337618282,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0923,
      "step": 37083
    },
    {
      "epoch": 8.827947390346962,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 37084
    },
    {
      "epoch": 8.828185443075641,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3782,
      "step": 37085
    },
    {
      "epoch": 8.828423495804321,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 37086
    },
    {
      "epoch": 8.828661548533,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0647,
      "step": 37087
    },
    {
      "epoch": 8.82889960126168,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 37088
    },
    {
      "epoch": 8.829137653990358,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.779,
      "step": 37089
    },
    {
      "epoch": 8.829375706719038,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 37090
    },
    {
      "epoch": 8.829613759447717,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0494,
      "step": 37091
    },
    {
      "epoch": 8.829851812176397,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2031,
      "step": 37092
    },
    {
      "epoch": 8.830089864905077,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.17,
      "step": 37093
    },
    {
      "epoch": 8.830327917633756,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1779,
      "step": 37094
    },
    {
      "epoch": 8.830565970362436,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0997,
      "step": 37095
    },
    {
      "epoch": 8.830804023091115,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3754,
      "step": 37096
    },
    {
      "epoch": 8.831042075819795,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 37097
    },
    {
      "epoch": 8.831280128548473,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 37098
    },
    {
      "epoch": 8.831518181277152,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0934,
      "step": 37099
    },
    {
      "epoch": 8.831756234005832,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.115,
      "step": 37100
    },
    {
      "epoch": 8.831756234005832,
      "eval_runtime": 0.5242,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 37100
    },
    {
      "epoch": 8.831994286734512,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 37101
    },
    {
      "epoch": 8.832232339463191,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 37102
    },
    {
      "epoch": 8.83247039219187,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1488,
      "step": 37103
    },
    {
      "epoch": 8.83270844492055,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 37104
    },
    {
      "epoch": 8.83294649764923,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2361,
      "step": 37105
    },
    {
      "epoch": 8.833184550377908,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0957,
      "step": 37106
    },
    {
      "epoch": 8.833422603106587,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0306,
      "step": 37107
    },
    {
      "epoch": 8.833660655835267,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 37108
    },
    {
      "epoch": 8.833898708563947,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 37109
    },
    {
      "epoch": 8.834136761292626,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 37110
    },
    {
      "epoch": 8.834374814021306,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.3258,
      "step": 37111
    },
    {
      "epoch": 8.834612866749985,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3524,
      "step": 37112
    },
    {
      "epoch": 8.834850919478665,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 37113
    },
    {
      "epoch": 8.835088972207345,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7471,
      "step": 37114
    },
    {
      "epoch": 8.835327024936023,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 37115
    },
    {
      "epoch": 8.835565077664702,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0563,
      "step": 37116
    },
    {
      "epoch": 8.835803130393382,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9966,
      "step": 37117
    },
    {
      "epoch": 8.836041183122061,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 37118
    },
    {
      "epoch": 8.836279235850741,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 37119
    },
    {
      "epoch": 8.83651728857942,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 37120
    },
    {
      "epoch": 8.8367553413081,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 37121
    },
    {
      "epoch": 8.83699339403678,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.218,
      "step": 37122
    },
    {
      "epoch": 8.83723144676546,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.2854,
      "step": 37123
    },
    {
      "epoch": 8.837469499494137,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9631,
      "step": 37124
    },
    {
      "epoch": 8.837707552222817,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 37125
    },
    {
      "epoch": 8.837945604951496,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 37126
    },
    {
      "epoch": 8.838183657680176,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9864,
      "step": 37127
    },
    {
      "epoch": 8.838421710408856,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 37128
    },
    {
      "epoch": 8.838659763137535,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 37129
    },
    {
      "epoch": 8.838897815866215,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 37130
    },
    {
      "epoch": 8.839135868594894,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.5082,
      "step": 37131
    },
    {
      "epoch": 8.839373921323574,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 37132
    },
    {
      "epoch": 8.839611974052252,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1907,
      "step": 37133
    },
    {
      "epoch": 8.839850026780931,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.217,
      "step": 37134
    },
    {
      "epoch": 8.840088079509611,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0581,
      "step": 37135
    },
    {
      "epoch": 8.84032613223829,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 37136
    },
    {
      "epoch": 8.84056418496697,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9887,
      "step": 37137
    },
    {
      "epoch": 8.84080223769565,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 37138
    },
    {
      "epoch": 8.84104029042433,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 37139
    },
    {
      "epoch": 8.841278343153009,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 37140
    },
    {
      "epoch": 8.841516395881687,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 37141
    },
    {
      "epoch": 8.841754448610367,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 37142
    },
    {
      "epoch": 8.841992501339046,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 37143
    },
    {
      "epoch": 8.842230554067726,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 37144
    },
    {
      "epoch": 8.842468606796405,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1811,
      "step": 37145
    },
    {
      "epoch": 8.842706659525085,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1437,
      "step": 37146
    },
    {
      "epoch": 8.842944712253765,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 37147
    },
    {
      "epoch": 8.843182764982444,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 37148
    },
    {
      "epoch": 8.843420817711124,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 37149
    },
    {
      "epoch": 8.843658870439802,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2598,
      "step": 37150
    },
    {
      "epoch": 8.843896923168481,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 37151
    },
    {
      "epoch": 8.84413497589716,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9453,
      "step": 37152
    },
    {
      "epoch": 8.84437302862584,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 37153
    },
    {
      "epoch": 8.84461108135452,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8967,
      "step": 37154
    },
    {
      "epoch": 8.8448491340832,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9789,
      "step": 37155
    },
    {
      "epoch": 8.84508718681188,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0892,
      "step": 37156
    },
    {
      "epoch": 8.845325239540559,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3667,
      "step": 37157
    },
    {
      "epoch": 8.845563292269238,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 37158
    },
    {
      "epoch": 8.845801344997916,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.4306,
      "step": 37159
    },
    {
      "epoch": 8.846039397726596,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1335,
      "step": 37160
    },
    {
      "epoch": 8.846277450455275,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 37161
    },
    {
      "epoch": 8.846515503183955,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1677,
      "step": 37162
    },
    {
      "epoch": 8.846753555912635,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.19,
      "step": 37163
    },
    {
      "epoch": 8.846991608641314,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1311,
      "step": 37164
    },
    {
      "epoch": 8.847229661369994,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 37165
    },
    {
      "epoch": 8.847467714098673,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3863,
      "step": 37166
    },
    {
      "epoch": 8.847705766827353,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9595,
      "step": 37167
    },
    {
      "epoch": 8.847943819556031,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2645,
      "step": 37168
    },
    {
      "epoch": 8.84818187228471,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 37169
    },
    {
      "epoch": 8.84841992501339,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 37170
    },
    {
      "epoch": 8.84865797774207,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.174,
      "step": 37171
    },
    {
      "epoch": 8.84889603047075,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1962,
      "step": 37172
    },
    {
      "epoch": 8.849134083199429,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9054,
      "step": 37173
    },
    {
      "epoch": 8.849372135928109,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0875,
      "step": 37174
    },
    {
      "epoch": 8.849610188656788,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 37175
    },
    {
      "epoch": 8.849848241385466,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1826,
      "step": 37176
    },
    {
      "epoch": 8.850086294114146,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0453,
      "step": 37177
    },
    {
      "epoch": 8.850324346842825,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 37178
    },
    {
      "epoch": 8.850562399571505,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 37179
    },
    {
      "epoch": 8.850800452300184,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 37180
    },
    {
      "epoch": 8.851038505028864,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9668,
      "step": 37181
    },
    {
      "epoch": 8.851276557757544,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9367,
      "step": 37182
    },
    {
      "epoch": 8.851514610486223,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 37183
    },
    {
      "epoch": 8.851752663214903,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1582,
      "step": 37184
    },
    {
      "epoch": 8.851990715943582,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7412,
      "step": 37185
    },
    {
      "epoch": 8.85222876867226,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 37186
    },
    {
      "epoch": 8.85246682140094,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.919,
      "step": 37187
    },
    {
      "epoch": 8.85270487412962,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 37188
    },
    {
      "epoch": 8.852942926858299,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 37189
    },
    {
      "epoch": 8.853180979586979,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 37190
    },
    {
      "epoch": 8.853419032315658,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 37191
    },
    {
      "epoch": 8.853657085044338,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 37192
    },
    {
      "epoch": 8.853895137773018,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 37193
    },
    {
      "epoch": 8.854133190501695,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9698,
      "step": 37194
    },
    {
      "epoch": 8.854371243230375,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 37195
    },
    {
      "epoch": 8.854609295959055,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 37196
    },
    {
      "epoch": 8.854847348687734,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 37197
    },
    {
      "epoch": 8.855085401416414,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 37198
    },
    {
      "epoch": 8.855323454145093,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2534,
      "step": 37199
    },
    {
      "epoch": 8.855561506873773,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 37200
    },
    {
      "epoch": 8.855561506873773,
      "eval_runtime": 0.5417,
      "eval_samples_per_second": 1.846,
      "eval_steps_per_second": 1.846,
      "step": 37200
    },
    {
      "epoch": 8.855799559602453,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 37201
    },
    {
      "epoch": 8.856037612331132,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 37202
    },
    {
      "epoch": 8.85627566505981,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 37203
    },
    {
      "epoch": 8.85651371778849,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1081,
      "step": 37204
    },
    {
      "epoch": 8.85675177051717,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 37205
    },
    {
      "epoch": 8.856989823245849,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1033,
      "step": 37206
    },
    {
      "epoch": 8.857227875974528,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1331,
      "step": 37207
    },
    {
      "epoch": 8.857465928703208,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 37208
    },
    {
      "epoch": 8.857703981431888,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0295,
      "step": 37209
    },
    {
      "epoch": 8.857942034160567,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1632,
      "step": 37210
    },
    {
      "epoch": 8.858180086889245,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 37211
    },
    {
      "epoch": 8.858418139617925,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0044,
      "step": 37212
    },
    {
      "epoch": 8.858656192346604,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0019,
      "step": 37213
    },
    {
      "epoch": 8.858894245075284,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1189,
      "step": 37214
    },
    {
      "epoch": 8.859132297803963,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1426,
      "step": 37215
    },
    {
      "epoch": 8.859370350532643,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 37216
    },
    {
      "epoch": 8.859608403261323,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 37217
    },
    {
      "epoch": 8.859846455990002,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 37218
    },
    {
      "epoch": 8.860084508718682,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 37219
    },
    {
      "epoch": 8.860322561447362,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2413,
      "step": 37220
    },
    {
      "epoch": 8.86056061417604,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.8515,
      "step": 37221
    },
    {
      "epoch": 8.860798666904719,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 37222
    },
    {
      "epoch": 8.861036719633399,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 37223
    },
    {
      "epoch": 8.861274772362078,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 37224
    },
    {
      "epoch": 8.861512825090758,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 37225
    },
    {
      "epoch": 8.861750877819437,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 37226
    },
    {
      "epoch": 8.861988930548117,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 2.5635,
      "step": 37227
    },
    {
      "epoch": 8.862226983276797,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1625,
      "step": 37228
    },
    {
      "epoch": 8.862465036005474,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.3009,
      "step": 37229
    },
    {
      "epoch": 8.862703088734154,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 37230
    },
    {
      "epoch": 8.862941141462834,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 37231
    },
    {
      "epoch": 8.863179194191513,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 37232
    },
    {
      "epoch": 8.863417246920193,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2989,
      "step": 37233
    },
    {
      "epoch": 8.863655299648872,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 37234
    },
    {
      "epoch": 8.863893352377552,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8918,
      "step": 37235
    },
    {
      "epoch": 8.864131405106232,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.4228,
      "step": 37236
    },
    {
      "epoch": 8.864369457834911,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7636,
      "step": 37237
    },
    {
      "epoch": 8.864607510563589,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 37238
    },
    {
      "epoch": 8.864845563292269,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 37239
    },
    {
      "epoch": 8.865083616020948,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 37240
    },
    {
      "epoch": 8.865321668749628,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1263,
      "step": 37241
    },
    {
      "epoch": 8.865559721478308,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 37242
    },
    {
      "epoch": 8.865797774206987,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 37243
    },
    {
      "epoch": 8.866035826935667,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9884,
      "step": 37244
    },
    {
      "epoch": 8.866273879664346,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.3188,
      "step": 37245
    },
    {
      "epoch": 8.866511932393024,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9681,
      "step": 37246
    },
    {
      "epoch": 8.866749985121704,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 37247
    },
    {
      "epoch": 8.866988037850383,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 37248
    },
    {
      "epoch": 8.867226090579063,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1259,
      "step": 37249
    },
    {
      "epoch": 8.867464143307743,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 37250
    },
    {
      "epoch": 8.867702196036422,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 37251
    },
    {
      "epoch": 8.867940248765102,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 37252
    },
    {
      "epoch": 8.868178301493781,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 37253
    },
    {
      "epoch": 8.868416354222461,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 37254
    },
    {
      "epoch": 8.86865440695114,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1369,
      "step": 37255
    },
    {
      "epoch": 8.868892459679818,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 37256
    },
    {
      "epoch": 8.869130512408498,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.172,
      "step": 37257
    },
    {
      "epoch": 8.869368565137178,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 37258
    },
    {
      "epoch": 8.869606617865857,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.4054,
      "step": 37259
    },
    {
      "epoch": 8.869844670594537,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 37260
    },
    {
      "epoch": 8.870082723323216,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 37261
    },
    {
      "epoch": 8.870320776051896,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 37262
    },
    {
      "epoch": 8.870558828780576,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2403,
      "step": 37263
    },
    {
      "epoch": 8.870796881509253,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 37264
    },
    {
      "epoch": 8.871034934237933,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 37265
    },
    {
      "epoch": 8.871272986966613,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1478,
      "step": 37266
    },
    {
      "epoch": 8.871511039695292,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8639,
      "step": 37267
    },
    {
      "epoch": 8.871749092423972,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8677,
      "step": 37268
    },
    {
      "epoch": 8.871987145152652,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 37269
    },
    {
      "epoch": 8.872225197881331,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0623,
      "step": 37270
    },
    {
      "epoch": 8.87246325061001,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1127,
      "step": 37271
    },
    {
      "epoch": 8.87270130333869,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2829,
      "step": 37272
    },
    {
      "epoch": 8.872939356067368,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8169,
      "step": 37273
    },
    {
      "epoch": 8.873177408796048,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 37274
    },
    {
      "epoch": 8.873415461524727,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 37275
    },
    {
      "epoch": 8.873653514253407,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 37276
    },
    {
      "epoch": 8.873891566982087,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 37277
    },
    {
      "epoch": 8.874129619710766,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 37278
    },
    {
      "epoch": 8.874367672439446,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8456,
      "step": 37279
    },
    {
      "epoch": 8.874605725168125,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.978,
      "step": 37280
    },
    {
      "epoch": 8.874843777896803,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 37281
    },
    {
      "epoch": 8.875081830625483,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1317,
      "step": 37282
    },
    {
      "epoch": 8.875319883354162,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1543,
      "step": 37283
    },
    {
      "epoch": 8.875557936082842,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 37284
    },
    {
      "epoch": 8.875795988811522,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 37285
    },
    {
      "epoch": 8.876034041540201,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 37286
    },
    {
      "epoch": 8.87627209426888,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7132,
      "step": 37287
    },
    {
      "epoch": 8.87651014699756,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 37288
    },
    {
      "epoch": 8.87674819972624,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2109,
      "step": 37289
    },
    {
      "epoch": 8.87698625245492,
      "grad_norm": 7.875,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 37290
    },
    {
      "epoch": 8.877224305183598,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0058,
      "step": 37291
    },
    {
      "epoch": 8.877462357912277,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 37292
    },
    {
      "epoch": 8.877700410640957,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7297,
      "step": 37293
    },
    {
      "epoch": 8.877938463369636,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 37294
    },
    {
      "epoch": 8.878176516098316,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 37295
    },
    {
      "epoch": 8.878414568826996,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 37296
    },
    {
      "epoch": 8.878652621555675,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.4296,
      "step": 37297
    },
    {
      "epoch": 8.878890674284355,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 37298
    },
    {
      "epoch": 8.879128727013033,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.8482,
      "step": 37299
    },
    {
      "epoch": 8.879366779741712,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 37300
    },
    {
      "epoch": 8.879366779741712,
      "eval_runtime": 0.5155,
      "eval_samples_per_second": 1.94,
      "eval_steps_per_second": 1.94,
      "step": 37300
    },
    {
      "epoch": 8.879604832470392,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 37301
    },
    {
      "epoch": 8.879842885199071,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.134,
      "step": 37302
    },
    {
      "epoch": 8.880080937927751,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 37303
    },
    {
      "epoch": 8.88031899065643,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7758,
      "step": 37304
    },
    {
      "epoch": 8.88055704338511,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 37305
    },
    {
      "epoch": 8.88079509611379,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0233,
      "step": 37306
    },
    {
      "epoch": 8.88103314884247,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 37307
    },
    {
      "epoch": 8.881271201571147,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 37308
    },
    {
      "epoch": 8.881509254299827,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 37309
    },
    {
      "epoch": 8.881747307028506,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1544,
      "step": 37310
    },
    {
      "epoch": 8.881985359757186,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 37311
    },
    {
      "epoch": 8.882223412485866,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 37312
    },
    {
      "epoch": 8.882461465214545,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8678,
      "step": 37313
    },
    {
      "epoch": 8.882699517943225,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 37314
    },
    {
      "epoch": 8.882937570671904,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1472,
      "step": 37315
    },
    {
      "epoch": 8.883175623400582,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 37316
    },
    {
      "epoch": 8.883413676129262,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 37317
    },
    {
      "epoch": 8.883651728857942,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 37318
    },
    {
      "epoch": 8.883889781586621,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9468,
      "step": 37319
    },
    {
      "epoch": 8.8841278343153,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1357,
      "step": 37320
    },
    {
      "epoch": 8.88436588704398,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 37321
    },
    {
      "epoch": 8.88460393977266,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1523,
      "step": 37322
    },
    {
      "epoch": 8.88484199250134,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 37323
    },
    {
      "epoch": 8.88508004523002,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1111,
      "step": 37324
    },
    {
      "epoch": 8.885318097958699,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0092,
      "step": 37325
    },
    {
      "epoch": 8.885556150687377,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 37326
    },
    {
      "epoch": 8.885794203416056,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2152,
      "step": 37327
    },
    {
      "epoch": 8.886032256144736,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0676,
      "step": 37328
    },
    {
      "epoch": 8.886270308873415,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8749,
      "step": 37329
    },
    {
      "epoch": 8.886508361602095,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0362,
      "step": 37330
    },
    {
      "epoch": 8.886746414330775,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0538,
      "step": 37331
    },
    {
      "epoch": 8.886984467059454,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 37332
    },
    {
      "epoch": 8.887222519788134,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 37333
    },
    {
      "epoch": 8.887460572516812,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8283,
      "step": 37334
    },
    {
      "epoch": 8.887698625245491,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 37335
    },
    {
      "epoch": 8.88793667797417,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0151,
      "step": 37336
    },
    {
      "epoch": 8.88817473070285,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 37337
    },
    {
      "epoch": 8.88841278343153,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0291,
      "step": 37338
    },
    {
      "epoch": 8.88865083616021,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 37339
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8755,
      "step": 37340
    },
    {
      "epoch": 8.889126941617569,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 37341
    },
    {
      "epoch": 8.889364994346248,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2985,
      "step": 37342
    },
    {
      "epoch": 8.889603047074926,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8937,
      "step": 37343
    },
    {
      "epoch": 8.889841099803606,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 37344
    },
    {
      "epoch": 8.890079152532286,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8348,
      "step": 37345
    },
    {
      "epoch": 8.890317205260965,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9721,
      "step": 37346
    },
    {
      "epoch": 8.890555257989645,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 37347
    },
    {
      "epoch": 8.890793310718324,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 37348
    },
    {
      "epoch": 8.891031363447004,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2998,
      "step": 37349
    },
    {
      "epoch": 8.891269416175684,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2222,
      "step": 37350
    },
    {
      "epoch": 8.891507468904361,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 37351
    },
    {
      "epoch": 8.891745521633041,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.0325,
      "step": 37352
    },
    {
      "epoch": 8.89198357436172,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 37353
    },
    {
      "epoch": 8.8922216270904,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 37354
    },
    {
      "epoch": 8.89245967981908,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 37355
    },
    {
      "epoch": 8.89269773254776,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 37356
    },
    {
      "epoch": 8.892935785276439,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9638,
      "step": 37357
    },
    {
      "epoch": 8.893173838005119,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1671,
      "step": 37358
    },
    {
      "epoch": 8.893411890733798,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0086,
      "step": 37359
    },
    {
      "epoch": 8.893649943462478,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 37360
    },
    {
      "epoch": 8.893887996191156,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 37361
    },
    {
      "epoch": 8.894126048919835,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 37362
    },
    {
      "epoch": 8.894364101648515,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 37363
    },
    {
      "epoch": 8.894602154377194,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1966,
      "step": 37364
    },
    {
      "epoch": 8.894840207105874,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 37365
    },
    {
      "epoch": 8.895078259834554,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 37366
    },
    {
      "epoch": 8.895316312563233,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 37367
    },
    {
      "epoch": 8.895554365291913,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0696,
      "step": 37368
    },
    {
      "epoch": 8.89579241802059,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1149,
      "step": 37369
    },
    {
      "epoch": 8.89603047074927,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 37370
    },
    {
      "epoch": 8.89626852347795,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2602,
      "step": 37371
    },
    {
      "epoch": 8.89650657620663,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 37372
    },
    {
      "epoch": 8.89674462893531,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 37373
    },
    {
      "epoch": 8.896982681663989,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 37374
    },
    {
      "epoch": 8.897220734392668,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8783,
      "step": 37375
    },
    {
      "epoch": 8.897458787121348,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2485,
      "step": 37376
    },
    {
      "epoch": 8.897696839850028,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 37377
    },
    {
      "epoch": 8.897934892578705,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1985,
      "step": 37378
    },
    {
      "epoch": 8.898172945307385,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 37379
    },
    {
      "epoch": 8.898410998036065,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9498,
      "step": 37380
    },
    {
      "epoch": 8.898649050764744,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8998,
      "step": 37381
    },
    {
      "epoch": 8.898887103493424,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8392,
      "step": 37382
    },
    {
      "epoch": 8.899125156222103,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1673,
      "step": 37383
    },
    {
      "epoch": 8.899363208950783,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.773,
      "step": 37384
    },
    {
      "epoch": 8.899601261679463,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1329,
      "step": 37385
    },
    {
      "epoch": 8.89983931440814,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0457,
      "step": 37386
    },
    {
      "epoch": 8.90007736713682,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9254,
      "step": 37387
    },
    {
      "epoch": 8.9003154198655,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1347,
      "step": 37388
    },
    {
      "epoch": 8.90055347259418,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0292,
      "step": 37389
    },
    {
      "epoch": 8.900791525322859,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 37390
    },
    {
      "epoch": 8.901029578051538,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 37391
    },
    {
      "epoch": 8.901267630780218,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.103,
      "step": 37392
    },
    {
      "epoch": 8.901505683508898,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 37393
    },
    {
      "epoch": 8.901743736237577,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 37394
    },
    {
      "epoch": 8.901981788966257,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 37395
    },
    {
      "epoch": 8.902219841694935,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7865,
      "step": 37396
    },
    {
      "epoch": 8.902457894423614,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 37397
    },
    {
      "epoch": 8.902695947152294,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9781,
      "step": 37398
    },
    {
      "epoch": 8.902933999880974,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 37399
    },
    {
      "epoch": 8.903172052609653,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1794,
      "step": 37400
    },
    {
      "epoch": 8.903172052609653,
      "eval_runtime": 0.5215,
      "eval_samples_per_second": 1.917,
      "eval_steps_per_second": 1.917,
      "step": 37400
    },
    {
      "epoch": 8.903410105338333,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8081,
      "step": 37401
    },
    {
      "epoch": 8.903648158067012,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 37402
    },
    {
      "epoch": 8.903886210795692,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 37403
    },
    {
      "epoch": 8.90412426352437,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 37404
    },
    {
      "epoch": 8.90436231625305,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 37405
    },
    {
      "epoch": 8.904600368981729,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 37406
    },
    {
      "epoch": 8.904838421710409,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 37407
    },
    {
      "epoch": 8.905076474439088,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9482,
      "step": 37408
    },
    {
      "epoch": 8.905314527167768,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9223,
      "step": 37409
    },
    {
      "epoch": 8.905552579896447,
      "grad_norm": 6.875,
      "learning_rate": 3e-05,
      "loss": 1.8861,
      "step": 37410
    },
    {
      "epoch": 8.905790632625127,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 37411
    },
    {
      "epoch": 8.906028685353807,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 37412
    },
    {
      "epoch": 8.906266738082484,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1463,
      "step": 37413
    },
    {
      "epoch": 8.906504790811164,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 37414
    },
    {
      "epoch": 8.906742843539844,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1051,
      "step": 37415
    },
    {
      "epoch": 8.906980896268523,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 37416
    },
    {
      "epoch": 8.907218948997203,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8303,
      "step": 37417
    },
    {
      "epoch": 8.907457001725883,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 37418
    },
    {
      "epoch": 8.907695054454562,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.6991,
      "step": 37419
    },
    {
      "epoch": 8.907933107183242,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1327,
      "step": 37420
    },
    {
      "epoch": 8.908171159911921,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0163,
      "step": 37421
    },
    {
      "epoch": 8.9084092126406,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.3056,
      "step": 37422
    },
    {
      "epoch": 8.908647265369279,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 37423
    },
    {
      "epoch": 8.908885318097958,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 37424
    },
    {
      "epoch": 8.909123370826638,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 37425
    },
    {
      "epoch": 8.909361423555318,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.954,
      "step": 37426
    },
    {
      "epoch": 8.909599476283997,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 37427
    },
    {
      "epoch": 8.909837529012677,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 37428
    },
    {
      "epoch": 8.910075581741356,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 37429
    },
    {
      "epoch": 8.910313634470036,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1564,
      "step": 37430
    },
    {
      "epoch": 8.910551687198714,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0419,
      "step": 37431
    },
    {
      "epoch": 8.910789739927393,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 37432
    },
    {
      "epoch": 8.911027792656073,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 37433
    },
    {
      "epoch": 8.911265845384753,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8142,
      "step": 37434
    },
    {
      "epoch": 8.911503898113432,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 37435
    },
    {
      "epoch": 8.911741950842112,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 37436
    },
    {
      "epoch": 8.911980003570791,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 37437
    },
    {
      "epoch": 8.912218056299471,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 37438
    },
    {
      "epoch": 8.912456109028149,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 37439
    },
    {
      "epoch": 8.912694161756828,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.927,
      "step": 37440
    },
    {
      "epoch": 8.912932214485508,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 37441
    },
    {
      "epoch": 8.913170267214188,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0559,
      "step": 37442
    },
    {
      "epoch": 8.913408319942867,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 37443
    },
    {
      "epoch": 8.913646372671547,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9531,
      "step": 37444
    },
    {
      "epoch": 8.913884425400227,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1587,
      "step": 37445
    },
    {
      "epoch": 8.914122478128906,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 37446
    },
    {
      "epoch": 8.914360530857586,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0231,
      "step": 37447
    },
    {
      "epoch": 8.914598583586264,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9775,
      "step": 37448
    },
    {
      "epoch": 8.914836636314943,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 37449
    },
    {
      "epoch": 8.915074689043623,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.22,
      "step": 37450
    },
    {
      "epoch": 8.915312741772302,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 37451
    },
    {
      "epoch": 8.915550794500982,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 37452
    },
    {
      "epoch": 8.915788847229662,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 37453
    },
    {
      "epoch": 8.916026899958341,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 37454
    },
    {
      "epoch": 8.91626495268702,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9664,
      "step": 37455
    },
    {
      "epoch": 8.9165030054157,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1943,
      "step": 37456
    },
    {
      "epoch": 8.916741058144378,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 37457
    },
    {
      "epoch": 8.916979110873058,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 37458
    },
    {
      "epoch": 8.917217163601737,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 37459
    },
    {
      "epoch": 8.917455216330417,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9309,
      "step": 37460
    },
    {
      "epoch": 8.917693269059097,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9187,
      "step": 37461
    },
    {
      "epoch": 8.917931321787776,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1162,
      "step": 37462
    },
    {
      "epoch": 8.918169374516456,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9882,
      "step": 37463
    },
    {
      "epoch": 8.918407427245135,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 37464
    },
    {
      "epoch": 8.918645479973815,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 37465
    },
    {
      "epoch": 8.918883532702493,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2553,
      "step": 37466
    },
    {
      "epoch": 8.919121585431173,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 37467
    },
    {
      "epoch": 8.919359638159852,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 37468
    },
    {
      "epoch": 8.919597690888532,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2664,
      "step": 37469
    },
    {
      "epoch": 8.919835743617211,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 37470
    },
    {
      "epoch": 8.920073796345891,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 37471
    },
    {
      "epoch": 8.92031184907457,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 37472
    },
    {
      "epoch": 8.92054990180325,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 37473
    },
    {
      "epoch": 8.920787954531928,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 37474
    },
    {
      "epoch": 8.921026007260608,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8422,
      "step": 37475
    },
    {
      "epoch": 8.921264059989287,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 37476
    },
    {
      "epoch": 8.921502112717967,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9509,
      "step": 37477
    },
    {
      "epoch": 8.921740165446646,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 37478
    },
    {
      "epoch": 8.921978218175326,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0077,
      "step": 37479
    },
    {
      "epoch": 8.922216270904006,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1115,
      "step": 37480
    },
    {
      "epoch": 8.922454323632685,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 37481
    },
    {
      "epoch": 8.922692376361365,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 37482
    },
    {
      "epoch": 8.922930429090043,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9192,
      "step": 37483
    },
    {
      "epoch": 8.923168481818722,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3916,
      "step": 37484
    },
    {
      "epoch": 8.923406534547402,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 37485
    },
    {
      "epoch": 8.923644587276081,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 37486
    },
    {
      "epoch": 8.923882640004761,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 37487
    },
    {
      "epoch": 8.92412069273344,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 37488
    },
    {
      "epoch": 8.92435874546212,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.105,
      "step": 37489
    },
    {
      "epoch": 8.9245967981908,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 37490
    },
    {
      "epoch": 8.92483485091948,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 37491
    },
    {
      "epoch": 8.925072903648157,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 37492
    },
    {
      "epoch": 8.925310956376837,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 37493
    },
    {
      "epoch": 8.925549009105517,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 37494
    },
    {
      "epoch": 8.925787061834196,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 37495
    },
    {
      "epoch": 8.926025114562876,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.183,
      "step": 37496
    },
    {
      "epoch": 8.926263167291555,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 37497
    },
    {
      "epoch": 8.926501220020235,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 37498
    },
    {
      "epoch": 8.926739272748915,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 37499
    },
    {
      "epoch": 8.926977325477594,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 37500
    },
    {
      "epoch": 8.926977325477594,
      "eval_runtime": 0.5195,
      "eval_samples_per_second": 1.925,
      "eval_steps_per_second": 1.925,
      "step": 37500
    },
    {
      "epoch": 8.927215378206272,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.312,
      "step": 37501
    },
    {
      "epoch": 8.927453430934952,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2201,
      "step": 37502
    },
    {
      "epoch": 8.927691483663631,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.3699,
      "step": 37503
    },
    {
      "epoch": 8.92792953639231,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8844,
      "step": 37504
    },
    {
      "epoch": 8.92816758912099,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0936,
      "step": 37505
    },
    {
      "epoch": 8.92840564184967,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 37506
    },
    {
      "epoch": 8.92864369457835,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 37507
    },
    {
      "epoch": 8.92888174730703,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 37508
    },
    {
      "epoch": 8.929119800035707,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9291,
      "step": 37509
    },
    {
      "epoch": 8.929357852764387,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 37510
    },
    {
      "epoch": 8.929595905493066,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1917,
      "step": 37511
    },
    {
      "epoch": 8.929833958221746,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8228,
      "step": 37512
    },
    {
      "epoch": 8.930072010950425,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 37513
    },
    {
      "epoch": 8.930310063679105,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8616,
      "step": 37514
    },
    {
      "epoch": 8.930548116407785,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 37515
    },
    {
      "epoch": 8.930786169136464,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8867,
      "step": 37516
    },
    {
      "epoch": 8.931024221865144,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 37517
    },
    {
      "epoch": 8.931262274593822,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 37518
    },
    {
      "epoch": 8.931500327322501,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 37519
    },
    {
      "epoch": 8.931738380051181,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0822,
      "step": 37520
    },
    {
      "epoch": 8.93197643277986,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8429,
      "step": 37521
    },
    {
      "epoch": 8.93221448550854,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 37522
    },
    {
      "epoch": 8.93245253823722,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1007,
      "step": 37523
    },
    {
      "epoch": 8.9326905909659,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 37524
    },
    {
      "epoch": 8.932928643694579,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 37525
    },
    {
      "epoch": 8.933166696423259,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1291,
      "step": 37526
    },
    {
      "epoch": 8.933404749151936,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8003,
      "step": 37527
    },
    {
      "epoch": 8.933642801880616,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 37528
    },
    {
      "epoch": 8.933880854609296,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0727,
      "step": 37529
    },
    {
      "epoch": 8.934118907337975,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9602,
      "step": 37530
    },
    {
      "epoch": 8.934356960066655,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 37531
    },
    {
      "epoch": 8.934595012795334,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 37532
    },
    {
      "epoch": 8.934833065524014,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1653,
      "step": 37533
    },
    {
      "epoch": 8.935071118252694,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 37534
    },
    {
      "epoch": 8.935309170981373,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 37535
    },
    {
      "epoch": 8.935547223710051,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.3186,
      "step": 37536
    },
    {
      "epoch": 8.93578527643873,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 37537
    },
    {
      "epoch": 8.93602332916741,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0541,
      "step": 37538
    },
    {
      "epoch": 8.93626138189609,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8539,
      "step": 37539
    },
    {
      "epoch": 8.93649943462477,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 37540
    },
    {
      "epoch": 8.936737487353449,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1557,
      "step": 37541
    },
    {
      "epoch": 8.936975540082129,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 37542
    },
    {
      "epoch": 8.937213592810808,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.6238,
      "step": 37543
    },
    {
      "epoch": 8.937451645539486,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9854,
      "step": 37544
    },
    {
      "epoch": 8.937689698268166,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.838,
      "step": 37545
    },
    {
      "epoch": 8.937927750996845,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1908,
      "step": 37546
    },
    {
      "epoch": 8.938165803725525,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 37547
    },
    {
      "epoch": 8.938403856454205,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6097,
      "step": 37548
    },
    {
      "epoch": 8.938641909182884,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 37549
    },
    {
      "epoch": 8.938879961911564,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 37550
    },
    {
      "epoch": 8.939118014640243,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 37551
    },
    {
      "epoch": 8.939356067368923,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9437,
      "step": 37552
    },
    {
      "epoch": 8.9395941200976,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 37553
    },
    {
      "epoch": 8.93983217282628,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9107,
      "step": 37554
    },
    {
      "epoch": 8.94007022555496,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 37555
    },
    {
      "epoch": 8.94030827828364,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 37556
    },
    {
      "epoch": 8.94054633101232,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 37557
    },
    {
      "epoch": 8.940784383740999,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 37558
    },
    {
      "epoch": 8.941022436469678,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0218,
      "step": 37559
    },
    {
      "epoch": 8.941260489198358,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 37560
    },
    {
      "epoch": 8.941498541927038,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.931,
      "step": 37561
    },
    {
      "epoch": 8.941736594655715,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0166,
      "step": 37562
    },
    {
      "epoch": 8.941974647384395,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0685,
      "step": 37563
    },
    {
      "epoch": 8.942212700113075,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.6379,
      "step": 37564
    },
    {
      "epoch": 8.942450752841754,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 37565
    },
    {
      "epoch": 8.942688805570434,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 37566
    },
    {
      "epoch": 8.942926858299113,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0134,
      "step": 37567
    },
    {
      "epoch": 8.943164911027793,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 37568
    },
    {
      "epoch": 8.943402963756473,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1511,
      "step": 37569
    },
    {
      "epoch": 8.943641016485152,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 37570
    },
    {
      "epoch": 8.94387906921383,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9603,
      "step": 37571
    },
    {
      "epoch": 8.94411712194251,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 37572
    },
    {
      "epoch": 8.94435517467119,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 37573
    },
    {
      "epoch": 8.944593227399869,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 37574
    },
    {
      "epoch": 8.944831280128549,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 37575
    },
    {
      "epoch": 8.945069332857228,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 37576
    },
    {
      "epoch": 8.945307385585908,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 37577
    },
    {
      "epoch": 8.945545438314587,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1138,
      "step": 37578
    },
    {
      "epoch": 8.945783491043265,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 37579
    },
    {
      "epoch": 8.946021543771945,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7893,
      "step": 37580
    },
    {
      "epoch": 8.946259596500624,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9827,
      "step": 37581
    },
    {
      "epoch": 8.946497649229304,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2581,
      "step": 37582
    },
    {
      "epoch": 8.946735701957984,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 37583
    },
    {
      "epoch": 8.946973754686663,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 37584
    },
    {
      "epoch": 8.947211807415343,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2151,
      "step": 37585
    },
    {
      "epoch": 8.947449860144022,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0651,
      "step": 37586
    },
    {
      "epoch": 8.947687912872702,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 37587
    },
    {
      "epoch": 8.947925965601382,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 37588
    },
    {
      "epoch": 8.94816401833006,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9996,
      "step": 37589
    },
    {
      "epoch": 8.948402071058739,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1004,
      "step": 37590
    },
    {
      "epoch": 8.948640123787419,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 37591
    },
    {
      "epoch": 8.948878176516098,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0046,
      "step": 37592
    },
    {
      "epoch": 8.949116229244778,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9018,
      "step": 37593
    },
    {
      "epoch": 8.949354281973458,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0618,
      "step": 37594
    },
    {
      "epoch": 8.949592334702137,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2962,
      "step": 37595
    },
    {
      "epoch": 8.949830387430817,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 37596
    },
    {
      "epoch": 8.950068440159495,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 37597
    },
    {
      "epoch": 8.950306492888174,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 37598
    },
    {
      "epoch": 8.950544545616854,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0728,
      "step": 37599
    },
    {
      "epoch": 8.950782598345533,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9675,
      "step": 37600
    },
    {
      "epoch": 8.950782598345533,
      "eval_runtime": 0.5181,
      "eval_samples_per_second": 1.93,
      "eval_steps_per_second": 1.93,
      "step": 37600
    },
    {
      "epoch": 8.951020651074213,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 37601
    },
    {
      "epoch": 8.951258703802893,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0302,
      "step": 37602
    },
    {
      "epoch": 8.951496756531572,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 37603
    },
    {
      "epoch": 8.951734809260252,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.3378,
      "step": 37604
    },
    {
      "epoch": 8.951972861988931,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 37605
    },
    {
      "epoch": 8.95221091471761,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 37606
    },
    {
      "epoch": 8.952448967446289,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 37607
    },
    {
      "epoch": 8.952687020174968,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9871,
      "step": 37608
    },
    {
      "epoch": 8.952925072903648,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3605,
      "step": 37609
    },
    {
      "epoch": 8.953163125632328,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 37610
    },
    {
      "epoch": 8.953401178361007,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 37611
    },
    {
      "epoch": 8.953639231089687,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 37612
    },
    {
      "epoch": 8.953877283818366,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 37613
    },
    {
      "epoch": 8.954115336547044,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 37614
    },
    {
      "epoch": 8.954353389275724,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0519,
      "step": 37615
    },
    {
      "epoch": 8.954591442004403,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 37616
    },
    {
      "epoch": 8.954829494733083,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8066,
      "step": 37617
    },
    {
      "epoch": 8.955067547461763,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.277,
      "step": 37618
    },
    {
      "epoch": 8.955305600190442,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 37619
    },
    {
      "epoch": 8.955543652919122,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 37620
    },
    {
      "epoch": 8.955781705647802,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 37621
    },
    {
      "epoch": 8.956019758376481,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 37622
    },
    {
      "epoch": 8.95625781110516,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0857,
      "step": 37623
    },
    {
      "epoch": 8.956495863833839,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2401,
      "step": 37624
    },
    {
      "epoch": 8.956733916562518,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1,
      "step": 37625
    },
    {
      "epoch": 8.956971969291198,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3262,
      "step": 37626
    },
    {
      "epoch": 8.957210022019877,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.067,
      "step": 37627
    },
    {
      "epoch": 8.957448074748557,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9191,
      "step": 37628
    },
    {
      "epoch": 8.957686127477237,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 1.8646,
      "step": 37629
    },
    {
      "epoch": 8.957924180205916,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2211,
      "step": 37630
    },
    {
      "epoch": 8.958162232934596,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.4621,
      "step": 37631
    },
    {
      "epoch": 8.958400285663274,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.3659,
      "step": 37632
    },
    {
      "epoch": 8.958638338391953,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1949,
      "step": 37633
    },
    {
      "epoch": 8.958876391120633,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 37634
    },
    {
      "epoch": 8.959114443849312,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 37635
    },
    {
      "epoch": 8.959352496577992,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1914,
      "step": 37636
    },
    {
      "epoch": 8.959590549306672,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.042,
      "step": 37637
    },
    {
      "epoch": 8.959828602035351,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 37638
    },
    {
      "epoch": 8.96006665476403,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 37639
    },
    {
      "epoch": 8.96030470749271,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1847,
      "step": 37640
    },
    {
      "epoch": 8.960542760221388,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 37641
    },
    {
      "epoch": 8.960780812950068,
      "grad_norm": 5.78125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 37642
    },
    {
      "epoch": 8.961018865678748,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 37643
    },
    {
      "epoch": 8.961256918407427,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 37644
    },
    {
      "epoch": 8.961494971136107,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8506,
      "step": 37645
    },
    {
      "epoch": 8.961733023864786,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 37646
    },
    {
      "epoch": 8.961971076593466,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9221,
      "step": 37647
    },
    {
      "epoch": 8.962209129322146,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 37648
    },
    {
      "epoch": 8.962447182050823,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 37649
    },
    {
      "epoch": 8.962685234779503,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8637,
      "step": 37650
    },
    {
      "epoch": 8.962923287508183,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 37651
    },
    {
      "epoch": 8.963161340236862,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 37652
    },
    {
      "epoch": 8.963399392965542,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8489,
      "step": 37653
    },
    {
      "epoch": 8.963637445694221,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0866,
      "step": 37654
    },
    {
      "epoch": 8.963875498422901,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 37655
    },
    {
      "epoch": 8.96411355115158,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 37656
    },
    {
      "epoch": 8.96435160388026,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 37657
    },
    {
      "epoch": 8.96458965660894,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8581,
      "step": 37658
    },
    {
      "epoch": 8.964827709337618,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1438,
      "step": 37659
    },
    {
      "epoch": 8.965065762066297,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 37660
    },
    {
      "epoch": 8.965303814794977,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8159,
      "step": 37661
    },
    {
      "epoch": 8.965541867523656,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0186,
      "step": 37662
    },
    {
      "epoch": 8.965779920252336,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0703,
      "step": 37663
    },
    {
      "epoch": 8.966017972981016,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 2.2637,
      "step": 37664
    },
    {
      "epoch": 8.966256025709695,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 37665
    },
    {
      "epoch": 8.966494078438375,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 37666
    },
    {
      "epoch": 8.966732131167053,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 37667
    },
    {
      "epoch": 8.966970183895732,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 37668
    },
    {
      "epoch": 8.967208236624412,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.383,
      "step": 37669
    },
    {
      "epoch": 8.967446289353092,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 37670
    },
    {
      "epoch": 8.967684342081771,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 37671
    },
    {
      "epoch": 8.96792239481045,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1076,
      "step": 37672
    },
    {
      "epoch": 8.96816044753913,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 37673
    },
    {
      "epoch": 8.96839850026781,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4126,
      "step": 37674
    },
    {
      "epoch": 8.96863655299649,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2046,
      "step": 37675
    },
    {
      "epoch": 8.968874605725167,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 37676
    },
    {
      "epoch": 8.969112658453847,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.313,
      "step": 37677
    },
    {
      "epoch": 8.969350711182527,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 37678
    },
    {
      "epoch": 8.969588763911206,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0564,
      "step": 37679
    },
    {
      "epoch": 8.969826816639886,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 37680
    },
    {
      "epoch": 8.970064869368565,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9744,
      "step": 37681
    },
    {
      "epoch": 8.970302922097245,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 37682
    },
    {
      "epoch": 8.970540974825925,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.3021,
      "step": 37683
    },
    {
      "epoch": 8.970779027554602,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1034,
      "step": 37684
    },
    {
      "epoch": 8.971017080283282,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 37685
    },
    {
      "epoch": 8.971255133011962,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 37686
    },
    {
      "epoch": 8.971493185740641,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 37687
    },
    {
      "epoch": 8.97173123846932,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2753,
      "step": 37688
    },
    {
      "epoch": 8.971969291198,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8114,
      "step": 37689
    },
    {
      "epoch": 8.97220734392668,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 37690
    },
    {
      "epoch": 8.97244539665536,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 37691
    },
    {
      "epoch": 8.97268344938404,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 37692
    },
    {
      "epoch": 8.972921502112719,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 37693
    },
    {
      "epoch": 8.973159554841397,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 37694
    },
    {
      "epoch": 8.973397607570076,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 37695
    },
    {
      "epoch": 8.973635660298756,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0511,
      "step": 37696
    },
    {
      "epoch": 8.973873713027436,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2493,
      "step": 37697
    },
    {
      "epoch": 8.974111765756115,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8672,
      "step": 37698
    },
    {
      "epoch": 8.974349818484795,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1652,
      "step": 37699
    },
    {
      "epoch": 8.974587871213474,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 37700
    },
    {
      "epoch": 8.974587871213474,
      "eval_runtime": 0.5263,
      "eval_samples_per_second": 1.9,
      "eval_steps_per_second": 1.9,
      "step": 37700
    },
    {
      "epoch": 8.974825923942154,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 37701
    },
    {
      "epoch": 8.975063976670832,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 37702
    },
    {
      "epoch": 8.975302029399511,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 37703
    },
    {
      "epoch": 8.975540082128191,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.6554,
      "step": 37704
    },
    {
      "epoch": 8.97577813485687,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1097,
      "step": 37705
    },
    {
      "epoch": 8.97601618758555,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 37706
    },
    {
      "epoch": 8.97625424031423,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 37707
    },
    {
      "epoch": 8.97649229304291,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 37708
    },
    {
      "epoch": 8.976730345771589,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 37709
    },
    {
      "epoch": 8.976968398500269,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 37710
    },
    {
      "epoch": 8.977206451228946,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7832,
      "step": 37711
    },
    {
      "epoch": 8.977444503957626,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 37712
    },
    {
      "epoch": 8.977682556686306,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 37713
    },
    {
      "epoch": 8.977920609414985,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1931,
      "step": 37714
    },
    {
      "epoch": 8.978158662143665,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 37715
    },
    {
      "epoch": 8.978396714872344,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9616,
      "step": 37716
    },
    {
      "epoch": 8.978634767601024,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0595,
      "step": 37717
    },
    {
      "epoch": 8.978872820329704,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 37718
    },
    {
      "epoch": 8.979110873058382,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1822,
      "step": 37719
    },
    {
      "epoch": 8.979348925787061,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 37720
    },
    {
      "epoch": 8.97958697851574,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 37721
    },
    {
      "epoch": 8.97982503124442,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 37722
    },
    {
      "epoch": 8.9800630839731,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 37723
    },
    {
      "epoch": 8.98030113670178,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1804,
      "step": 37724
    },
    {
      "epoch": 8.98053918943046,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 37725
    },
    {
      "epoch": 8.980777242159139,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 37726
    },
    {
      "epoch": 8.981015294887818,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.5425,
      "step": 37727
    },
    {
      "epoch": 8.981253347616498,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7977,
      "step": 37728
    },
    {
      "epoch": 8.981491400345176,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.5835,
      "step": 37729
    },
    {
      "epoch": 8.981729453073855,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 37730
    },
    {
      "epoch": 8.981967505802535,
      "grad_norm": 5.96875,
      "learning_rate": 3e-05,
      "loss": 1.9013,
      "step": 37731
    },
    {
      "epoch": 8.982205558531215,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0072,
      "step": 37732
    },
    {
      "epoch": 8.982443611259894,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 37733
    },
    {
      "epoch": 8.982681663988574,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.234,
      "step": 37734
    },
    {
      "epoch": 8.982919716717253,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 37735
    },
    {
      "epoch": 8.983157769445933,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.776,
      "step": 37736
    },
    {
      "epoch": 8.98339582217461,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0156,
      "step": 37737
    },
    {
      "epoch": 8.98363387490329,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2288,
      "step": 37738
    },
    {
      "epoch": 8.98387192763197,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.056,
      "step": 37739
    },
    {
      "epoch": 8.98410998036065,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1353,
      "step": 37740
    },
    {
      "epoch": 8.98434803308933,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 37741
    },
    {
      "epoch": 8.984586085818009,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1184,
      "step": 37742
    },
    {
      "epoch": 8.984824138546688,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 37743
    },
    {
      "epoch": 8.985062191275368,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 37744
    },
    {
      "epoch": 8.985300244004048,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8736,
      "step": 37745
    },
    {
      "epoch": 8.985538296732726,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 37746
    },
    {
      "epoch": 8.985776349461405,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.044,
      "step": 37747
    },
    {
      "epoch": 8.986014402190085,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2458,
      "step": 37748
    },
    {
      "epoch": 8.986252454918764,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8876,
      "step": 37749
    },
    {
      "epoch": 8.986490507647444,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9305,
      "step": 37750
    },
    {
      "epoch": 8.986728560376124,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 37751
    },
    {
      "epoch": 8.986966613104803,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2634,
      "step": 37752
    },
    {
      "epoch": 8.987204665833483,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 37753
    },
    {
      "epoch": 8.98744271856216,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 37754
    },
    {
      "epoch": 8.98768077129084,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6334,
      "step": 37755
    },
    {
      "epoch": 8.98791882401952,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 37756
    },
    {
      "epoch": 8.9881568767482,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 1.9782,
      "step": 37757
    },
    {
      "epoch": 8.988394929476879,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2102,
      "step": 37758
    },
    {
      "epoch": 8.988632982205559,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9255,
      "step": 37759
    },
    {
      "epoch": 8.988871034934238,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9417,
      "step": 37760
    },
    {
      "epoch": 8.989109087662918,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 37761
    },
    {
      "epoch": 8.989347140391597,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 37762
    },
    {
      "epoch": 8.989585193120277,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 37763
    },
    {
      "epoch": 8.989823245848955,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 37764
    },
    {
      "epoch": 8.990061298577634,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 37765
    },
    {
      "epoch": 8.990299351306314,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.144,
      "step": 37766
    },
    {
      "epoch": 8.990537404034994,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2738,
      "step": 37767
    },
    {
      "epoch": 8.990775456763673,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 37768
    },
    {
      "epoch": 8.991013509492353,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 37769
    },
    {
      "epoch": 8.991251562221033,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1515,
      "step": 37770
    },
    {
      "epoch": 8.991489614949712,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 37771
    },
    {
      "epoch": 8.99172766767839,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 37772
    },
    {
      "epoch": 8.99196572040707,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7445,
      "step": 37773
    },
    {
      "epoch": 8.99220377313575,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8791,
      "step": 37774
    },
    {
      "epoch": 8.992441825864429,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 37775
    },
    {
      "epoch": 8.992679878593108,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.2597,
      "step": 37776
    },
    {
      "epoch": 8.992917931321788,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1075,
      "step": 37777
    },
    {
      "epoch": 8.993155984050468,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1941,
      "step": 37778
    },
    {
      "epoch": 8.993394036779147,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 37779
    },
    {
      "epoch": 8.993632089507827,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 37780
    },
    {
      "epoch": 8.993870142236505,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9501,
      "step": 37781
    },
    {
      "epoch": 8.994108194965184,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2111,
      "step": 37782
    },
    {
      "epoch": 8.994346247693864,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0767,
      "step": 37783
    },
    {
      "epoch": 8.994584300422543,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 37784
    },
    {
      "epoch": 8.994822353151223,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8325,
      "step": 37785
    },
    {
      "epoch": 8.995060405879903,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 37786
    },
    {
      "epoch": 8.995298458608582,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1925,
      "step": 37787
    },
    {
      "epoch": 8.995536511337262,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 37788
    },
    {
      "epoch": 8.99577456406594,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0317,
      "step": 37789
    },
    {
      "epoch": 8.99601261679462,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0608,
      "step": 37790
    },
    {
      "epoch": 8.996250669523299,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1055,
      "step": 37791
    },
    {
      "epoch": 8.996488722251978,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 37792
    },
    {
      "epoch": 8.996726774980658,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 37793
    },
    {
      "epoch": 8.996964827709338,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9803,
      "step": 37794
    },
    {
      "epoch": 8.997202880438017,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 37795
    },
    {
      "epoch": 8.997440933166697,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1685,
      "step": 37796
    },
    {
      "epoch": 8.997678985895377,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 37797
    },
    {
      "epoch": 8.997917038624056,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1539,
      "step": 37798
    },
    {
      "epoch": 8.998155091352734,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9848,
      "step": 37799
    },
    {
      "epoch": 8.998393144081414,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0386,
      "step": 37800
    },
    {
      "epoch": 8.998393144081414,
      "eval_runtime": 0.5171,
      "eval_samples_per_second": 1.934,
      "eval_steps_per_second": 1.934,
      "step": 37800
    },
    {
      "epoch": 8.998631196810093,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 37801
    },
    {
      "epoch": 8.998869249538773,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7784,
      "step": 37802
    },
    {
      "epoch": 8.999107302267452,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 37803
    },
    {
      "epoch": 8.999345354996132,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.073,
      "step": 37804
    },
    {
      "epoch": 8.999583407724812,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 37805
    },
    {
      "epoch": 8.999821460453491,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1417,
      "step": 37806
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1074,
      "step": 37807
    },
    {
      "epoch": 9.00023805272868,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 37808
    },
    {
      "epoch": 9.00047610545736,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8827,
      "step": 37809
    },
    {
      "epoch": 9.000714158186039,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 37810
    },
    {
      "epoch": 9.000952210914718,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 37811
    },
    {
      "epoch": 9.001190263643396,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 37812
    },
    {
      "epoch": 9.001428316372076,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 37813
    },
    {
      "epoch": 9.001666369100755,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.6888,
      "step": 37814
    },
    {
      "epoch": 9.001904421829435,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.271,
      "step": 37815
    },
    {
      "epoch": 9.002142474558115,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 37816
    },
    {
      "epoch": 9.002380527286794,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 37817
    },
    {
      "epoch": 9.002618580015474,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2775,
      "step": 37818
    },
    {
      "epoch": 9.002856632744153,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 37819
    },
    {
      "epoch": 9.003094685472833,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 37820
    },
    {
      "epoch": 9.003332738201511,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 37821
    },
    {
      "epoch": 9.00357079093019,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 37822
    },
    {
      "epoch": 9.00380884365887,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2348,
      "step": 37823
    },
    {
      "epoch": 9.00404689638755,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2104,
      "step": 37824
    },
    {
      "epoch": 9.00428494911623,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 37825
    },
    {
      "epoch": 9.004523001844909,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9327,
      "step": 37826
    },
    {
      "epoch": 9.004761054573589,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0235,
      "step": 37827
    },
    {
      "epoch": 9.004999107302268,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 37828
    },
    {
      "epoch": 9.005237160030946,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.7609,
      "step": 37829
    },
    {
      "epoch": 9.005475212759626,
      "grad_norm": 3.015625,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 37830
    },
    {
      "epoch": 9.005713265488305,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8617,
      "step": 37831
    },
    {
      "epoch": 9.005951318216985,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.292,
      "step": 37832
    },
    {
      "epoch": 9.006189370945664,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0569,
      "step": 37833
    },
    {
      "epoch": 9.006427423674344,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7999,
      "step": 37834
    },
    {
      "epoch": 9.006665476403024,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 37835
    },
    {
      "epoch": 9.006903529131703,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 37836
    },
    {
      "epoch": 9.007141581860383,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9943,
      "step": 37837
    },
    {
      "epoch": 9.00737963458906,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 37838
    },
    {
      "epoch": 9.00761768731774,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2617,
      "step": 37839
    },
    {
      "epoch": 9.00785574004642,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8526,
      "step": 37840
    },
    {
      "epoch": 9.0080937927751,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7363,
      "step": 37841
    },
    {
      "epoch": 9.008331845503779,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8049,
      "step": 37842
    },
    {
      "epoch": 9.008569898232459,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.837,
      "step": 37843
    },
    {
      "epoch": 9.008807950961138,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 37844
    },
    {
      "epoch": 9.009046003689818,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7344,
      "step": 37845
    },
    {
      "epoch": 9.009284056418497,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 37846
    },
    {
      "epoch": 9.009522109147175,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 37847
    },
    {
      "epoch": 9.009760161875855,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 37848
    },
    {
      "epoch": 9.009998214604535,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 37849
    },
    {
      "epoch": 9.010236267333214,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 37850
    },
    {
      "epoch": 9.010474320061894,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2654,
      "step": 37851
    },
    {
      "epoch": 9.010712372790573,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2804,
      "step": 37852
    },
    {
      "epoch": 9.010950425519253,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 37853
    },
    {
      "epoch": 9.011188478247933,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.032,
      "step": 37854
    },
    {
      "epoch": 9.011426530976612,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0312,
      "step": 37855
    },
    {
      "epoch": 9.01166458370529,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 37856
    },
    {
      "epoch": 9.01190263643397,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 37857
    },
    {
      "epoch": 9.01214068916265,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9461,
      "step": 37858
    },
    {
      "epoch": 9.012378741891329,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 37859
    },
    {
      "epoch": 9.012616794620008,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 37860
    },
    {
      "epoch": 9.012854847348688,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.084,
      "step": 37861
    },
    {
      "epoch": 9.013092900077368,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.4456,
      "step": 37862
    },
    {
      "epoch": 9.013330952806047,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 37863
    },
    {
      "epoch": 9.013569005534725,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 37864
    },
    {
      "epoch": 9.013807058263405,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0429,
      "step": 37865
    },
    {
      "epoch": 9.014045110992084,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 37866
    },
    {
      "epoch": 9.014283163720764,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 37867
    },
    {
      "epoch": 9.014521216449443,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 37868
    },
    {
      "epoch": 9.014759269178123,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.156,
      "step": 37869
    },
    {
      "epoch": 9.014997321906803,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 37870
    },
    {
      "epoch": 9.015235374635482,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0768,
      "step": 37871
    },
    {
      "epoch": 9.015473427364162,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7881,
      "step": 37872
    },
    {
      "epoch": 9.01571148009284,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 37873
    },
    {
      "epoch": 9.01594953282152,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 37874
    },
    {
      "epoch": 9.016187585550199,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0807,
      "step": 37875
    },
    {
      "epoch": 9.016425638278879,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 37876
    },
    {
      "epoch": 9.016663691007558,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1136,
      "step": 37877
    },
    {
      "epoch": 9.016901743736238,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2932,
      "step": 37878
    },
    {
      "epoch": 9.017139796464917,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1754,
      "step": 37879
    },
    {
      "epoch": 9.017377849193597,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.211,
      "step": 37880
    },
    {
      "epoch": 9.017615901922277,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 37881
    },
    {
      "epoch": 9.017853954650954,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 37882
    },
    {
      "epoch": 9.018092007379634,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.968,
      "step": 37883
    },
    {
      "epoch": 9.018330060108314,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 37884
    },
    {
      "epoch": 9.018568112836993,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 37885
    },
    {
      "epoch": 9.018806165565673,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 37886
    },
    {
      "epoch": 9.019044218294352,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9084,
      "step": 37887
    },
    {
      "epoch": 9.019282271023032,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1001,
      "step": 37888
    },
    {
      "epoch": 9.019520323751712,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 37889
    },
    {
      "epoch": 9.019758376480391,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8001,
      "step": 37890
    },
    {
      "epoch": 9.019996429209069,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7583,
      "step": 37891
    },
    {
      "epoch": 9.020234481937749,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8959,
      "step": 37892
    },
    {
      "epoch": 9.020472534666428,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.9627,
      "step": 37893
    },
    {
      "epoch": 9.020710587395108,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1476,
      "step": 37894
    },
    {
      "epoch": 9.020948640123787,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 37895
    },
    {
      "epoch": 9.021186692852467,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9236,
      "step": 37896
    },
    {
      "epoch": 9.021424745581147,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9452,
      "step": 37897
    },
    {
      "epoch": 9.021662798309826,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 37898
    },
    {
      "epoch": 9.021900851038504,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7619,
      "step": 37899
    },
    {
      "epoch": 9.022138903767184,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1986,
      "step": 37900
    },
    {
      "epoch": 9.022138903767184,
      "eval_runtime": 0.513,
      "eval_samples_per_second": 1.949,
      "eval_steps_per_second": 1.949,
      "step": 37900
    },
    {
      "epoch": 9.022376956495863,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 37901
    },
    {
      "epoch": 9.022615009224543,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8084,
      "step": 37902
    },
    {
      "epoch": 9.022853061953223,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.4565,
      "step": 37903
    },
    {
      "epoch": 9.023091114681902,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 37904
    },
    {
      "epoch": 9.023329167410582,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 37905
    },
    {
      "epoch": 9.023567220139261,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 37906
    },
    {
      "epoch": 9.023805272867941,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9059,
      "step": 37907
    },
    {
      "epoch": 9.024043325596619,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 37908
    },
    {
      "epoch": 9.024281378325298,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 37909
    },
    {
      "epoch": 9.024519431053978,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 37910
    },
    {
      "epoch": 9.024757483782658,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9777,
      "step": 37911
    },
    {
      "epoch": 9.024995536511337,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2908,
      "step": 37912
    },
    {
      "epoch": 9.025233589240017,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9728,
      "step": 37913
    },
    {
      "epoch": 9.025471641968696,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 37914
    },
    {
      "epoch": 9.025709694697376,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 37915
    },
    {
      "epoch": 9.025947747426056,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 37916
    },
    {
      "epoch": 9.026185800154733,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0706,
      "step": 37917
    },
    {
      "epoch": 9.026423852883413,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2019,
      "step": 37918
    },
    {
      "epoch": 9.026661905612093,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2817,
      "step": 37919
    },
    {
      "epoch": 9.026899958340772,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9386,
      "step": 37920
    },
    {
      "epoch": 9.027138011069452,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 37921
    },
    {
      "epoch": 9.027376063798132,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0238,
      "step": 37922
    },
    {
      "epoch": 9.027614116526811,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0637,
      "step": 37923
    },
    {
      "epoch": 9.02785216925549,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 37924
    },
    {
      "epoch": 9.02809022198417,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 37925
    },
    {
      "epoch": 9.028328274712848,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 37926
    },
    {
      "epoch": 9.028566327441528,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.8567,
      "step": 37927
    },
    {
      "epoch": 9.028804380170207,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 37928
    },
    {
      "epoch": 9.029042432898887,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2444,
      "step": 37929
    },
    {
      "epoch": 9.029280485627567,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.6761,
      "step": 37930
    },
    {
      "epoch": 9.029518538356246,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 37931
    },
    {
      "epoch": 9.029756591084926,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9053,
      "step": 37932
    },
    {
      "epoch": 9.029994643813605,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.5303,
      "step": 37933
    },
    {
      "epoch": 9.030232696542285,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8439,
      "step": 37934
    },
    {
      "epoch": 9.030470749270963,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 37935
    },
    {
      "epoch": 9.030708801999642,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7309,
      "step": 37936
    },
    {
      "epoch": 9.030946854728322,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.112,
      "step": 37937
    },
    {
      "epoch": 9.031184907457002,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1842,
      "step": 37938
    },
    {
      "epoch": 9.031422960185681,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7672,
      "step": 37939
    },
    {
      "epoch": 9.03166101291436,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 37940
    },
    {
      "epoch": 9.03189906564304,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 37941
    },
    {
      "epoch": 9.03213711837172,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 37942
    },
    {
      "epoch": 9.032375171100398,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 37943
    },
    {
      "epoch": 9.032613223829077,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 37944
    },
    {
      "epoch": 9.032851276557757,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1497,
      "step": 37945
    },
    {
      "epoch": 9.033089329286437,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 37946
    },
    {
      "epoch": 9.033327382015116,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1188,
      "step": 37947
    },
    {
      "epoch": 9.033565434743796,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 37948
    },
    {
      "epoch": 9.033803487472476,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7354,
      "step": 37949
    },
    {
      "epoch": 9.034041540201155,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8592,
      "step": 37950
    },
    {
      "epoch": 9.034279592929835,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2171,
      "step": 37951
    },
    {
      "epoch": 9.034517645658513,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8898,
      "step": 37952
    },
    {
      "epoch": 9.034755698387192,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1348,
      "step": 37953
    },
    {
      "epoch": 9.034993751115872,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0502,
      "step": 37954
    },
    {
      "epoch": 9.035231803844551,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 37955
    },
    {
      "epoch": 9.035469856573231,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1242,
      "step": 37956
    },
    {
      "epoch": 9.03570790930191,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 37957
    },
    {
      "epoch": 9.03594596203059,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8147,
      "step": 37958
    },
    {
      "epoch": 9.03618401475927,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.6934,
      "step": 37959
    },
    {
      "epoch": 9.03642206748795,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 37960
    },
    {
      "epoch": 9.036660120216627,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9745,
      "step": 37961
    },
    {
      "epoch": 9.036898172945307,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8018,
      "step": 37962
    },
    {
      "epoch": 9.037136225673986,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 37963
    },
    {
      "epoch": 9.037374278402666,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8889,
      "step": 37964
    },
    {
      "epoch": 9.037612331131346,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.3551,
      "step": 37965
    },
    {
      "epoch": 9.037850383860025,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.3026,
      "step": 37966
    },
    {
      "epoch": 9.038088436588705,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 37967
    },
    {
      "epoch": 9.038326489317384,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 37968
    },
    {
      "epoch": 9.038564542046064,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1289,
      "step": 37969
    },
    {
      "epoch": 9.038802594774742,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 37970
    },
    {
      "epoch": 9.039040647503422,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.898,
      "step": 37971
    },
    {
      "epoch": 9.039278700232101,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 37972
    },
    {
      "epoch": 9.03951675296078,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9209,
      "step": 37973
    },
    {
      "epoch": 9.03975480568946,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 37974
    },
    {
      "epoch": 9.03999285841814,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3574,
      "step": 37975
    },
    {
      "epoch": 9.04023091114682,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 37976
    },
    {
      "epoch": 9.0404689638755,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 37977
    },
    {
      "epoch": 9.040707016604177,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 37978
    },
    {
      "epoch": 9.040945069332857,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0417,
      "step": 37979
    },
    {
      "epoch": 9.041183122061536,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 37980
    },
    {
      "epoch": 9.041421174790216,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 37981
    },
    {
      "epoch": 9.041659227518895,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2442,
      "step": 37982
    },
    {
      "epoch": 9.041897280247575,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 37983
    },
    {
      "epoch": 9.042135332976255,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 37984
    },
    {
      "epoch": 9.042373385704934,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.5019,
      "step": 37985
    },
    {
      "epoch": 9.042611438433614,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.3784,
      "step": 37986
    },
    {
      "epoch": 9.042849491162292,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 37987
    },
    {
      "epoch": 9.043087543890971,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 37988
    },
    {
      "epoch": 9.04332559661965,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.041,
      "step": 37989
    },
    {
      "epoch": 9.04356364934833,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 37990
    },
    {
      "epoch": 9.04380170207701,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9338,
      "step": 37991
    },
    {
      "epoch": 9.04403975480569,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0982,
      "step": 37992
    },
    {
      "epoch": 9.04427780753437,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0074,
      "step": 37993
    },
    {
      "epoch": 9.044515860263049,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.2381,
      "step": 37994
    },
    {
      "epoch": 9.044753912991728,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 37995
    },
    {
      "epoch": 9.044991965720406,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.092,
      "step": 37996
    },
    {
      "epoch": 9.045230018449086,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.8839,
      "step": 37997
    },
    {
      "epoch": 9.045468071177766,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 1.7125,
      "step": 37998
    },
    {
      "epoch": 9.045706123906445,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4215,
      "step": 37999
    },
    {
      "epoch": 9.045944176635125,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 38000
    },
    {
      "epoch": 9.045944176635125,
      "eval_runtime": 0.5145,
      "eval_samples_per_second": 1.944,
      "eval_steps_per_second": 1.944,
      "step": 38000
    },
    {
      "epoch": 9.046182229363804,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 1.7578,
      "step": 38001
    },
    {
      "epoch": 9.046420282092484,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9971,
      "step": 38002
    },
    {
      "epoch": 9.046658334821164,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8943,
      "step": 38003
    },
    {
      "epoch": 9.046896387549843,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 38004
    },
    {
      "epoch": 9.047134440278521,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 38005
    },
    {
      "epoch": 9.0473724930072,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.83,
      "step": 38006
    },
    {
      "epoch": 9.04761054573588,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.264,
      "step": 38007
    },
    {
      "epoch": 9.04784859846456,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 38008
    },
    {
      "epoch": 9.04808665119324,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 38009
    },
    {
      "epoch": 9.048324703921919,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 38010
    },
    {
      "epoch": 9.048562756650599,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 38011
    },
    {
      "epoch": 9.048800809379278,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 38012
    },
    {
      "epoch": 9.049038862107956,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8353,
      "step": 38013
    },
    {
      "epoch": 9.049276914836636,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.297,
      "step": 38014
    },
    {
      "epoch": 9.049514967565315,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2499,
      "step": 38015
    },
    {
      "epoch": 9.049753020293995,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3394,
      "step": 38016
    },
    {
      "epoch": 9.049991073022674,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9809,
      "step": 38017
    },
    {
      "epoch": 9.050229125751354,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 38018
    },
    {
      "epoch": 9.050467178480034,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 38019
    },
    {
      "epoch": 9.050705231208713,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8522,
      "step": 38020
    },
    {
      "epoch": 9.050943283937393,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9926,
      "step": 38021
    },
    {
      "epoch": 9.05118133666607,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 38022
    },
    {
      "epoch": 9.05141938939475,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 38023
    },
    {
      "epoch": 9.05165744212343,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 38024
    },
    {
      "epoch": 9.05189549485211,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 38025
    },
    {
      "epoch": 9.05213354758079,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 38026
    },
    {
      "epoch": 9.052371600309469,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 38027
    },
    {
      "epoch": 9.052609653038148,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2725,
      "step": 38028
    },
    {
      "epoch": 9.052847705766828,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9256,
      "step": 38029
    },
    {
      "epoch": 9.053085758495508,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 38030
    },
    {
      "epoch": 9.053323811224185,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 38031
    },
    {
      "epoch": 9.053561863952865,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 38032
    },
    {
      "epoch": 9.053799916681545,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 38033
    },
    {
      "epoch": 9.054037969410224,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 38034
    },
    {
      "epoch": 9.054276022138904,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 38035
    },
    {
      "epoch": 9.054514074867583,
      "grad_norm": 6.53125,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 38036
    },
    {
      "epoch": 9.054752127596263,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 38037
    },
    {
      "epoch": 9.054990180324943,
      "grad_norm": 6.9375,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 38038
    },
    {
      "epoch": 9.055228233053622,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.9614,
      "step": 38039
    },
    {
      "epoch": 9.0554662857823,
      "grad_norm": 6.25,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 38040
    },
    {
      "epoch": 9.05570433851098,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 38041
    },
    {
      "epoch": 9.05594239123966,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1393,
      "step": 38042
    },
    {
      "epoch": 9.056180443968339,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 38043
    },
    {
      "epoch": 9.056418496697018,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.0968,
      "step": 38044
    },
    {
      "epoch": 9.056656549425698,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 38045
    },
    {
      "epoch": 9.056894602154378,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9793,
      "step": 38046
    },
    {
      "epoch": 9.057132654883057,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8963,
      "step": 38047
    },
    {
      "epoch": 9.057370707611735,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8914,
      "step": 38048
    },
    {
      "epoch": 9.057608760340415,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 38049
    },
    {
      "epoch": 9.057846813069094,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 38050
    },
    {
      "epoch": 9.058084865797774,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0407,
      "step": 38051
    },
    {
      "epoch": 9.058322918526454,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 38052
    },
    {
      "epoch": 9.058560971255133,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0763,
      "step": 38053
    },
    {
      "epoch": 9.058799023983813,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 38054
    },
    {
      "epoch": 9.059037076712492,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 38055
    },
    {
      "epoch": 9.059275129441172,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.5935,
      "step": 38056
    },
    {
      "epoch": 9.05951318216985,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2106,
      "step": 38057
    },
    {
      "epoch": 9.05975123489853,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 38058
    },
    {
      "epoch": 9.059989287627209,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 38059
    },
    {
      "epoch": 9.060227340355889,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 38060
    },
    {
      "epoch": 9.060465393084568,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 38061
    },
    {
      "epoch": 9.060703445813248,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8284,
      "step": 38062
    },
    {
      "epoch": 9.060941498541927,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1787,
      "step": 38063
    },
    {
      "epoch": 9.061179551270607,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9139,
      "step": 38064
    },
    {
      "epoch": 9.061417603999287,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 38065
    },
    {
      "epoch": 9.061655656727964,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 38066
    },
    {
      "epoch": 9.061893709456644,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 38067
    },
    {
      "epoch": 9.062131762185324,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 38068
    },
    {
      "epoch": 9.062369814914003,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 38069
    },
    {
      "epoch": 9.062607867642683,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6895,
      "step": 38070
    },
    {
      "epoch": 9.062845920371362,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8232,
      "step": 38071
    },
    {
      "epoch": 9.063083973100042,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7723,
      "step": 38072
    },
    {
      "epoch": 9.063322025828722,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 38073
    },
    {
      "epoch": 9.063560078557401,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 38074
    },
    {
      "epoch": 9.06379813128608,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 38075
    },
    {
      "epoch": 9.064036184014759,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 38076
    },
    {
      "epoch": 9.064274236743438,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 38077
    },
    {
      "epoch": 9.064512289472118,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7417,
      "step": 38078
    },
    {
      "epoch": 9.064750342200798,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 38079
    },
    {
      "epoch": 9.064988394929477,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2178,
      "step": 38080
    },
    {
      "epoch": 9.065226447658157,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0658,
      "step": 38081
    },
    {
      "epoch": 9.065464500386836,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 38082
    },
    {
      "epoch": 9.065702553115514,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 38083
    },
    {
      "epoch": 9.065940605844194,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 38084
    },
    {
      "epoch": 9.066178658572873,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0164,
      "step": 38085
    },
    {
      "epoch": 9.066416711301553,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 38086
    },
    {
      "epoch": 9.066654764030233,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 38087
    },
    {
      "epoch": 9.066892816758912,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 38088
    },
    {
      "epoch": 9.067130869487592,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2311,
      "step": 38089
    },
    {
      "epoch": 9.067368922216271,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0198,
      "step": 38090
    },
    {
      "epoch": 9.067606974944951,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4051,
      "step": 38091
    },
    {
      "epoch": 9.067845027673629,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 38092
    },
    {
      "epoch": 9.068083080402308,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 38093
    },
    {
      "epoch": 9.068321133130988,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8665,
      "step": 38094
    },
    {
      "epoch": 9.068559185859668,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.202,
      "step": 38095
    },
    {
      "epoch": 9.068797238588347,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 38096
    },
    {
      "epoch": 9.069035291317027,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0723,
      "step": 38097
    },
    {
      "epoch": 9.069273344045707,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 38098
    },
    {
      "epoch": 9.069511396774386,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.3492,
      "step": 38099
    },
    {
      "epoch": 9.069749449503066,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.141,
      "step": 38100
    },
    {
      "epoch": 9.069749449503066,
      "eval_runtime": 0.511,
      "eval_samples_per_second": 1.957,
      "eval_steps_per_second": 1.957,
      "step": 38100
    },
    {
      "epoch": 9.069987502231744,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.7311,
      "step": 38101
    },
    {
      "epoch": 9.070225554960423,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 38102
    },
    {
      "epoch": 9.070463607689103,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0397,
      "step": 38103
    },
    {
      "epoch": 9.070701660417782,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 38104
    },
    {
      "epoch": 9.070939713146462,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8582,
      "step": 38105
    },
    {
      "epoch": 9.071177765875142,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 38106
    },
    {
      "epoch": 9.071415818603821,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2205,
      "step": 38107
    },
    {
      "epoch": 9.0716538713325,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 38108
    },
    {
      "epoch": 9.07189192406118,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.0088,
      "step": 38109
    },
    {
      "epoch": 9.072129976789858,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 38110
    },
    {
      "epoch": 9.072368029518538,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.2343,
      "step": 38111
    },
    {
      "epoch": 9.072606082247217,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 38112
    },
    {
      "epoch": 9.072844134975897,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8207,
      "step": 38113
    },
    {
      "epoch": 9.073082187704577,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7779,
      "step": 38114
    },
    {
      "epoch": 9.073320240433256,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 38115
    },
    {
      "epoch": 9.073558293161936,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 38116
    },
    {
      "epoch": 9.073796345890615,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 38117
    },
    {
      "epoch": 9.074034398619293,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 38118
    },
    {
      "epoch": 9.074272451347973,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2087,
      "step": 38119
    },
    {
      "epoch": 9.074510504076652,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 38120
    },
    {
      "epoch": 9.074748556805332,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9695,
      "step": 38121
    },
    {
      "epoch": 9.074986609534012,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 38122
    },
    {
      "epoch": 9.075224662262691,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8671,
      "step": 38123
    },
    {
      "epoch": 9.075462714991371,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 38124
    },
    {
      "epoch": 9.07570076772005,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8162,
      "step": 38125
    },
    {
      "epoch": 9.07593882044873,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.02,
      "step": 38126
    },
    {
      "epoch": 9.076176873177408,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 38127
    },
    {
      "epoch": 9.076414925906088,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 38128
    },
    {
      "epoch": 9.076652978634767,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2214,
      "step": 38129
    },
    {
      "epoch": 9.076891031363447,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 38130
    },
    {
      "epoch": 9.077129084092126,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 38131
    },
    {
      "epoch": 9.077367136820806,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 38132
    },
    {
      "epoch": 9.077605189549486,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.924,
      "step": 38133
    },
    {
      "epoch": 9.077843242278165,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0458,
      "step": 38134
    },
    {
      "epoch": 9.078081295006845,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3005,
      "step": 38135
    },
    {
      "epoch": 9.078319347735523,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 38136
    },
    {
      "epoch": 9.078557400464202,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1355,
      "step": 38137
    },
    {
      "epoch": 9.078795453192882,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 38138
    },
    {
      "epoch": 9.079033505921561,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 38139
    },
    {
      "epoch": 9.079271558650241,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.216,
      "step": 38140
    },
    {
      "epoch": 9.07950961137892,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 38141
    },
    {
      "epoch": 9.0797476641076,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0816,
      "step": 38142
    },
    {
      "epoch": 9.07998571683628,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 38143
    },
    {
      "epoch": 9.08022376956496,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9147,
      "step": 38144
    },
    {
      "epoch": 9.080461822293637,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0695,
      "step": 38145
    },
    {
      "epoch": 9.080699875022317,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 38146
    },
    {
      "epoch": 9.080937927750997,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 38147
    },
    {
      "epoch": 9.081175980479676,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 38148
    },
    {
      "epoch": 9.081414033208356,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 38149
    },
    {
      "epoch": 9.081652085937035,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9704,
      "step": 38150
    },
    {
      "epoch": 9.081890138665715,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 38151
    },
    {
      "epoch": 9.082128191394395,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 38152
    },
    {
      "epoch": 9.082366244123074,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 38153
    },
    {
      "epoch": 9.082604296851752,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0732,
      "step": 38154
    },
    {
      "epoch": 9.082842349580432,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2882,
      "step": 38155
    },
    {
      "epoch": 9.083080402309111,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7302,
      "step": 38156
    },
    {
      "epoch": 9.08331845503779,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1891,
      "step": 38157
    },
    {
      "epoch": 9.08355650776647,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0862,
      "step": 38158
    },
    {
      "epoch": 9.08379456049515,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 38159
    },
    {
      "epoch": 9.08403261322383,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 38160
    },
    {
      "epoch": 9.08427066595251,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3116,
      "step": 38161
    },
    {
      "epoch": 9.084508718681187,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.034,
      "step": 38162
    },
    {
      "epoch": 9.084746771409867,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0666,
      "step": 38163
    },
    {
      "epoch": 9.084984824138546,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 38164
    },
    {
      "epoch": 9.085222876867226,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2944,
      "step": 38165
    },
    {
      "epoch": 9.085460929595905,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0808,
      "step": 38166
    },
    {
      "epoch": 9.085698982324585,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9592,
      "step": 38167
    },
    {
      "epoch": 9.085937035053265,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7405,
      "step": 38168
    },
    {
      "epoch": 9.086175087781944,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.932,
      "step": 38169
    },
    {
      "epoch": 9.086413140510624,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0179,
      "step": 38170
    },
    {
      "epoch": 9.086651193239302,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9122,
      "step": 38171
    },
    {
      "epoch": 9.086889245967981,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0902,
      "step": 38172
    },
    {
      "epoch": 9.087127298696661,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8801,
      "step": 38173
    },
    {
      "epoch": 9.08736535142534,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.962,
      "step": 38174
    },
    {
      "epoch": 9.08760340415402,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 38175
    },
    {
      "epoch": 9.0878414568827,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 38176
    },
    {
      "epoch": 9.08807950961138,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 38177
    },
    {
      "epoch": 9.088317562340059,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.421,
      "step": 38178
    },
    {
      "epoch": 9.088555615068739,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 38179
    },
    {
      "epoch": 9.088793667797416,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 38180
    },
    {
      "epoch": 9.089031720526096,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0912,
      "step": 38181
    },
    {
      "epoch": 9.089269773254776,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7347,
      "step": 38182
    },
    {
      "epoch": 9.089507825983455,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 38183
    },
    {
      "epoch": 9.089745878712135,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.058,
      "step": 38184
    },
    {
      "epoch": 9.089983931440814,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2188,
      "step": 38185
    },
    {
      "epoch": 9.090221984169494,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 38186
    },
    {
      "epoch": 9.090460036898174,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2402,
      "step": 38187
    },
    {
      "epoch": 9.090698089626853,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 38188
    },
    {
      "epoch": 9.090936142355531,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8759,
      "step": 38189
    },
    {
      "epoch": 9.09117419508421,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8878,
      "step": 38190
    },
    {
      "epoch": 9.09141224781289,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0006,
      "step": 38191
    },
    {
      "epoch": 9.09165030054157,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8537,
      "step": 38192
    },
    {
      "epoch": 9.09188835327025,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.3456,
      "step": 38193
    },
    {
      "epoch": 9.092126405998929,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0047,
      "step": 38194
    },
    {
      "epoch": 9.092364458727609,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 38195
    },
    {
      "epoch": 9.092602511456288,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0361,
      "step": 38196
    },
    {
      "epoch": 9.092840564184966,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 38197
    },
    {
      "epoch": 9.093078616913646,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8023,
      "step": 38198
    },
    {
      "epoch": 9.093316669642325,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.133,
      "step": 38199
    },
    {
      "epoch": 9.093554722371005,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9175,
      "step": 38200
    },
    {
      "epoch": 9.093554722371005,
      "eval_runtime": 0.5296,
      "eval_samples_per_second": 1.888,
      "eval_steps_per_second": 1.888,
      "step": 38200
    },
    {
      "epoch": 9.093792775099685,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 38201
    },
    {
      "epoch": 9.094030827828364,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 38202
    },
    {
      "epoch": 9.094268880557044,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9561,
      "step": 38203
    },
    {
      "epoch": 9.094506933285723,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 38204
    },
    {
      "epoch": 9.094744986014403,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.257,
      "step": 38205
    },
    {
      "epoch": 9.09498303874308,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 38206
    },
    {
      "epoch": 9.09522109147176,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0477,
      "step": 38207
    },
    {
      "epoch": 9.09545914420044,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0288,
      "step": 38208
    },
    {
      "epoch": 9.09569719692912,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 38209
    },
    {
      "epoch": 9.0959352496578,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.3732,
      "step": 38210
    },
    {
      "epoch": 9.096173302386479,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.844,
      "step": 38211
    },
    {
      "epoch": 9.096411355115158,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1246,
      "step": 38212
    },
    {
      "epoch": 9.096649407843838,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6693,
      "step": 38213
    },
    {
      "epoch": 9.096887460572518,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 38214
    },
    {
      "epoch": 9.097125513301195,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 38215
    },
    {
      "epoch": 9.097363566029875,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 38216
    },
    {
      "epoch": 9.097601618758555,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9802,
      "step": 38217
    },
    {
      "epoch": 9.097839671487234,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 38218
    },
    {
      "epoch": 9.098077724215914,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 38219
    },
    {
      "epoch": 9.098315776944593,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9705,
      "step": 38220
    },
    {
      "epoch": 9.098553829673273,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.3882,
      "step": 38221
    },
    {
      "epoch": 9.098791882401953,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 38222
    },
    {
      "epoch": 9.099029935130632,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0911,
      "step": 38223
    },
    {
      "epoch": 9.09926798785931,
      "grad_norm": 5.8125,
      "learning_rate": 3e-05,
      "loss": 2.3622,
      "step": 38224
    },
    {
      "epoch": 9.09950604058799,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 38225
    },
    {
      "epoch": 9.09974409331667,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 38226
    },
    {
      "epoch": 9.099982146045349,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 38227
    },
    {
      "epoch": 9.100220198774029,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9972,
      "step": 38228
    },
    {
      "epoch": 9.100458251502708,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1328,
      "step": 38229
    },
    {
      "epoch": 9.100696304231388,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9123,
      "step": 38230
    },
    {
      "epoch": 9.100934356960067,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.2359,
      "step": 38231
    },
    {
      "epoch": 9.101172409688745,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 38232
    },
    {
      "epoch": 9.101410462417425,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 38233
    },
    {
      "epoch": 9.101648515146104,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 38234
    },
    {
      "epoch": 9.101886567874784,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2318,
      "step": 38235
    },
    {
      "epoch": 9.102124620603464,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 38236
    },
    {
      "epoch": 9.102362673332143,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8993,
      "step": 38237
    },
    {
      "epoch": 9.102600726060823,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.839,
      "step": 38238
    },
    {
      "epoch": 9.102838778789502,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1827,
      "step": 38239
    },
    {
      "epoch": 9.103076831518182,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 38240
    },
    {
      "epoch": 9.10331488424686,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 38241
    },
    {
      "epoch": 9.10355293697554,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1765,
      "step": 38242
    },
    {
      "epoch": 9.103790989704219,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 38243
    },
    {
      "epoch": 9.104029042432899,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0304,
      "step": 38244
    },
    {
      "epoch": 9.104267095161578,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 38245
    },
    {
      "epoch": 9.104505147890258,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9353,
      "step": 38246
    },
    {
      "epoch": 9.104743200618937,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 38247
    },
    {
      "epoch": 9.104981253347617,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 38248
    },
    {
      "epoch": 9.105219306076297,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2314,
      "step": 38249
    },
    {
      "epoch": 9.105457358804975,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 38250
    },
    {
      "epoch": 9.105695411533654,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 38251
    },
    {
      "epoch": 9.105933464262334,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 38252
    },
    {
      "epoch": 9.106171516991013,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1359,
      "step": 38253
    },
    {
      "epoch": 9.106409569719693,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 38254
    },
    {
      "epoch": 9.106647622448373,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 38255
    },
    {
      "epoch": 9.106885675177052,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0961,
      "step": 38256
    },
    {
      "epoch": 9.107123727905732,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 38257
    },
    {
      "epoch": 9.107361780634411,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1781,
      "step": 38258
    },
    {
      "epoch": 9.10759983336309,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 38259
    },
    {
      "epoch": 9.107837886091769,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0185,
      "step": 38260
    },
    {
      "epoch": 9.108075938820448,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8072,
      "step": 38261
    },
    {
      "epoch": 9.108313991549128,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 38262
    },
    {
      "epoch": 9.108552044277808,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9702,
      "step": 38263
    },
    {
      "epoch": 9.108790097006487,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9768,
      "step": 38264
    },
    {
      "epoch": 9.109028149735167,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8723,
      "step": 38265
    },
    {
      "epoch": 9.109266202463846,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 38266
    },
    {
      "epoch": 9.109504255192524,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 38267
    },
    {
      "epoch": 9.109742307921204,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 38268
    },
    {
      "epoch": 9.109980360649883,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 38269
    },
    {
      "epoch": 9.110218413378563,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2107,
      "step": 38270
    },
    {
      "epoch": 9.110456466107243,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2158,
      "step": 38271
    },
    {
      "epoch": 9.110694518835922,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 38272
    },
    {
      "epoch": 9.110932571564602,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9195,
      "step": 38273
    },
    {
      "epoch": 9.111170624293282,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 38274
    },
    {
      "epoch": 9.111408677021961,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2076,
      "step": 38275
    },
    {
      "epoch": 9.111646729750639,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0645,
      "step": 38276
    },
    {
      "epoch": 9.111884782479319,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9503,
      "step": 38277
    },
    {
      "epoch": 9.112122835207998,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 38278
    },
    {
      "epoch": 9.112360887936678,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0757,
      "step": 38279
    },
    {
      "epoch": 9.112598940665357,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9565,
      "step": 38280
    },
    {
      "epoch": 9.112836993394037,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 38281
    },
    {
      "epoch": 9.113075046122717,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 38282
    },
    {
      "epoch": 9.113313098851396,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3199,
      "step": 38283
    },
    {
      "epoch": 9.113551151580076,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0776,
      "step": 38284
    },
    {
      "epoch": 9.113789204308754,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.064,
      "step": 38285
    },
    {
      "epoch": 9.114027257037433,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.3016,
      "step": 38286
    },
    {
      "epoch": 9.114265309766113,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9376,
      "step": 38287
    },
    {
      "epoch": 9.114503362494792,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9944,
      "step": 38288
    },
    {
      "epoch": 9.114741415223472,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 38289
    },
    {
      "epoch": 9.114979467952152,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9697,
      "step": 38290
    },
    {
      "epoch": 9.115217520680831,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 38291
    },
    {
      "epoch": 9.11545557340951,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 38292
    },
    {
      "epoch": 9.11569362613819,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1023,
      "step": 38293
    },
    {
      "epoch": 9.115931678866868,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9962,
      "step": 38294
    },
    {
      "epoch": 9.116169731595548,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 38295
    },
    {
      "epoch": 9.116407784324227,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 38296
    },
    {
      "epoch": 9.116645837052907,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 38297
    },
    {
      "epoch": 9.116883889781587,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7947,
      "step": 38298
    },
    {
      "epoch": 9.117121942510266,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 38299
    },
    {
      "epoch": 9.117359995238946,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0995,
      "step": 38300
    },
    {
      "epoch": 9.117359995238946,
      "eval_runtime": 0.5197,
      "eval_samples_per_second": 1.924,
      "eval_steps_per_second": 1.924,
      "step": 38300
    },
    {
      "epoch": 9.117598047967626,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0689,
      "step": 38301
    },
    {
      "epoch": 9.117836100696305,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 38302
    },
    {
      "epoch": 9.118074153424983,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0369,
      "step": 38303
    },
    {
      "epoch": 9.118312206153663,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 38304
    },
    {
      "epoch": 9.118550258882342,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 38305
    },
    {
      "epoch": 9.118788311611022,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8578,
      "step": 38306
    },
    {
      "epoch": 9.119026364339701,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 38307
    },
    {
      "epoch": 9.119264417068381,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 38308
    },
    {
      "epoch": 9.11950246979706,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 38309
    },
    {
      "epoch": 9.11974052252574,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1303,
      "step": 38310
    },
    {
      "epoch": 9.119978575254418,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 38311
    },
    {
      "epoch": 9.120216627983098,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2963,
      "step": 38312
    },
    {
      "epoch": 9.120454680711777,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 38313
    },
    {
      "epoch": 9.120692733440457,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 38314
    },
    {
      "epoch": 9.120930786169136,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 38315
    },
    {
      "epoch": 9.121168838897816,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.699,
      "step": 38316
    },
    {
      "epoch": 9.121406891626496,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.222,
      "step": 38317
    },
    {
      "epoch": 9.121644944355175,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0097,
      "step": 38318
    },
    {
      "epoch": 9.121882997083855,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.057,
      "step": 38319
    },
    {
      "epoch": 9.122121049812533,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2284,
      "step": 38320
    },
    {
      "epoch": 9.122359102541212,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 38321
    },
    {
      "epoch": 9.122597155269892,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.06,
      "step": 38322
    },
    {
      "epoch": 9.122835207998572,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.5788,
      "step": 38323
    },
    {
      "epoch": 9.123073260727251,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0008,
      "step": 38324
    },
    {
      "epoch": 9.12331131345593,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 38325
    },
    {
      "epoch": 9.12354936618461,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1316,
      "step": 38326
    },
    {
      "epoch": 9.12378741891329,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.695,
      "step": 38327
    },
    {
      "epoch": 9.12402547164197,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 38328
    },
    {
      "epoch": 9.124263524370647,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.2856,
      "step": 38329
    },
    {
      "epoch": 9.124501577099327,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 2.1284,
      "step": 38330
    },
    {
      "epoch": 9.124739629828007,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 38331
    },
    {
      "epoch": 9.124977682556686,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.021,
      "step": 38332
    },
    {
      "epoch": 9.125215735285366,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 38333
    },
    {
      "epoch": 9.125453788014045,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 38334
    },
    {
      "epoch": 9.125691840742725,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 38335
    },
    {
      "epoch": 9.125929893471405,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9644,
      "step": 38336
    },
    {
      "epoch": 9.126167946200084,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 38337
    },
    {
      "epoch": 9.126405998928762,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 38338
    },
    {
      "epoch": 9.126644051657442,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 38339
    },
    {
      "epoch": 9.126882104386121,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2012,
      "step": 38340
    },
    {
      "epoch": 9.1271201571148,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1864,
      "step": 38341
    },
    {
      "epoch": 9.12735820984348,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 38342
    },
    {
      "epoch": 9.12759626257216,
      "grad_norm": 6.0,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 38343
    },
    {
      "epoch": 9.12783431530084,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 38344
    },
    {
      "epoch": 9.12807236802952,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2063,
      "step": 38345
    },
    {
      "epoch": 9.128310420758197,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 38346
    },
    {
      "epoch": 9.128548473486877,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.9976,
      "step": 38347
    },
    {
      "epoch": 9.128786526215556,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 38348
    },
    {
      "epoch": 9.129024578944236,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2397,
      "step": 38349
    },
    {
      "epoch": 9.129262631672916,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3565,
      "step": 38350
    },
    {
      "epoch": 9.129500684401595,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 38351
    },
    {
      "epoch": 9.129738737130275,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2181,
      "step": 38352
    },
    {
      "epoch": 9.129976789858954,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1818,
      "step": 38353
    },
    {
      "epoch": 9.130214842587634,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1107,
      "step": 38354
    },
    {
      "epoch": 9.130452895316312,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 38355
    },
    {
      "epoch": 9.130690948044991,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 38356
    },
    {
      "epoch": 9.130929000773671,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.94,
      "step": 38357
    },
    {
      "epoch": 9.13116705350235,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.7134,
      "step": 38358
    },
    {
      "epoch": 9.13140510623103,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 38359
    },
    {
      "epoch": 9.13164315895971,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1385,
      "step": 38360
    },
    {
      "epoch": 9.13188121168839,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.4731,
      "step": 38361
    },
    {
      "epoch": 9.132119264417069,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9914,
      "step": 38362
    },
    {
      "epoch": 9.132357317145749,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 38363
    },
    {
      "epoch": 9.132595369874426,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1486,
      "step": 38364
    },
    {
      "epoch": 9.132833422603106,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0461,
      "step": 38365
    },
    {
      "epoch": 9.133071475331786,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9999,
      "step": 38366
    },
    {
      "epoch": 9.133309528060465,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 38367
    },
    {
      "epoch": 9.133547580789145,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7046,
      "step": 38368
    },
    {
      "epoch": 9.133785633517824,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 38369
    },
    {
      "epoch": 9.134023686246504,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1786,
      "step": 38370
    },
    {
      "epoch": 9.134261738975184,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 38371
    },
    {
      "epoch": 9.134499791703863,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1446,
      "step": 38372
    },
    {
      "epoch": 9.134737844432541,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9443,
      "step": 38373
    },
    {
      "epoch": 9.13497589716122,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2248,
      "step": 38374
    },
    {
      "epoch": 9.1352139498899,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 38375
    },
    {
      "epoch": 9.13545200261858,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 38376
    },
    {
      "epoch": 9.13569005534726,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 38377
    },
    {
      "epoch": 9.13592810807594,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 38378
    },
    {
      "epoch": 9.136166160804619,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2833,
      "step": 38379
    },
    {
      "epoch": 9.136404213533298,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9797,
      "step": 38380
    },
    {
      "epoch": 9.136642266261976,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.153,
      "step": 38381
    },
    {
      "epoch": 9.136880318990656,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 38382
    },
    {
      "epoch": 9.137118371719335,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 38383
    },
    {
      "epoch": 9.137356424448015,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 38384
    },
    {
      "epoch": 9.137594477176695,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 38385
    },
    {
      "epoch": 9.137832529905374,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.049,
      "step": 38386
    },
    {
      "epoch": 9.138070582634054,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 38387
    },
    {
      "epoch": 9.138308635362733,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 38388
    },
    {
      "epoch": 9.138546688091413,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.016,
      "step": 38389
    },
    {
      "epoch": 9.13878474082009,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8948,
      "step": 38390
    },
    {
      "epoch": 9.13902279354877,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.7906,
      "step": 38391
    },
    {
      "epoch": 9.13926084627745,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2944,
      "step": 38392
    },
    {
      "epoch": 9.13949889900613,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8803,
      "step": 38393
    },
    {
      "epoch": 9.13973695173481,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8832,
      "step": 38394
    },
    {
      "epoch": 9.139975004463489,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2189,
      "step": 38395
    },
    {
      "epoch": 9.140213057192168,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.2508,
      "step": 38396
    },
    {
      "epoch": 9.140451109920848,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 38397
    },
    {
      "epoch": 9.140689162649528,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 38398
    },
    {
      "epoch": 9.140927215378206,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 38399
    },
    {
      "epoch": 9.141165268106885,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.828,
      "step": 38400
    },
    {
      "epoch": 9.141165268106885,
      "eval_runtime": 0.5139,
      "eval_samples_per_second": 1.946,
      "eval_steps_per_second": 1.946,
      "step": 38400
    },
    {
      "epoch": 9.141403320835565,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2682,
      "step": 38401
    },
    {
      "epoch": 9.141641373564244,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3218,
      "step": 38402
    },
    {
      "epoch": 9.141879426292924,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.182,
      "step": 38403
    },
    {
      "epoch": 9.142117479021604,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 38404
    },
    {
      "epoch": 9.142355531750283,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2172,
      "step": 38405
    },
    {
      "epoch": 9.142593584478963,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9629,
      "step": 38406
    },
    {
      "epoch": 9.142831637207642,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2721,
      "step": 38407
    },
    {
      "epoch": 9.14306968993632,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.321,
      "step": 38408
    },
    {
      "epoch": 9.143307742665,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 38409
    },
    {
      "epoch": 9.14354579539368,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 38410
    },
    {
      "epoch": 9.143783848122359,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.442,
      "step": 38411
    },
    {
      "epoch": 9.144021900851039,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 38412
    },
    {
      "epoch": 9.144259953579718,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 38413
    },
    {
      "epoch": 9.144498006308398,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1723,
      "step": 38414
    },
    {
      "epoch": 9.144736059037077,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 38415
    },
    {
      "epoch": 9.144974111765755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1852,
      "step": 38416
    },
    {
      "epoch": 9.145212164494435,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 38417
    },
    {
      "epoch": 9.145450217223114,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 38418
    },
    {
      "epoch": 9.145688269951794,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.6994,
      "step": 38419
    },
    {
      "epoch": 9.145926322680474,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 38420
    },
    {
      "epoch": 9.146164375409153,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 38421
    },
    {
      "epoch": 9.146402428137833,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8399,
      "step": 38422
    },
    {
      "epoch": 9.146640480866512,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9206,
      "step": 38423
    },
    {
      "epoch": 9.146878533595192,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 38424
    },
    {
      "epoch": 9.14711658632387,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1342,
      "step": 38425
    },
    {
      "epoch": 9.14735463905255,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 38426
    },
    {
      "epoch": 9.14759269178123,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8992,
      "step": 38427
    },
    {
      "epoch": 9.147830744509909,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 38428
    },
    {
      "epoch": 9.148068797238588,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 38429
    },
    {
      "epoch": 9.148306849967268,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 38430
    },
    {
      "epoch": 9.148544902695948,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9774,
      "step": 38431
    },
    {
      "epoch": 9.148782955424627,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8936,
      "step": 38432
    },
    {
      "epoch": 9.149021008153307,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 38433
    },
    {
      "epoch": 9.149259060881985,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 38434
    },
    {
      "epoch": 9.149497113610664,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 38435
    },
    {
      "epoch": 9.149735166339344,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.129,
      "step": 38436
    },
    {
      "epoch": 9.149973219068023,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 38437
    },
    {
      "epoch": 9.150211271796703,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8535,
      "step": 38438
    },
    {
      "epoch": 9.150449324525383,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 38439
    },
    {
      "epoch": 9.150687377254062,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 38440
    },
    {
      "epoch": 9.150925429982742,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 38441
    },
    {
      "epoch": 9.151163482711421,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2468,
      "step": 38442
    },
    {
      "epoch": 9.1514015354401,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.6654,
      "step": 38443
    },
    {
      "epoch": 9.151639588168779,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 38444
    },
    {
      "epoch": 9.151877640897458,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 38445
    },
    {
      "epoch": 9.152115693626138,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 38446
    },
    {
      "epoch": 9.152353746354818,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 38447
    },
    {
      "epoch": 9.152591799083497,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 38448
    },
    {
      "epoch": 9.152829851812177,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 38449
    },
    {
      "epoch": 9.153067904540857,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9335,
      "step": 38450
    },
    {
      "epoch": 9.153305957269534,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 38451
    },
    {
      "epoch": 9.153544009998214,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9663,
      "step": 38452
    },
    {
      "epoch": 9.153782062726894,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 38453
    },
    {
      "epoch": 9.154020115455573,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1727,
      "step": 38454
    },
    {
      "epoch": 9.154258168184253,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9411,
      "step": 38455
    },
    {
      "epoch": 9.154496220912932,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 38456
    },
    {
      "epoch": 9.154734273641612,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 38457
    },
    {
      "epoch": 9.154972326370292,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 38458
    },
    {
      "epoch": 9.155210379098971,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 38459
    },
    {
      "epoch": 9.155448431827649,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1445,
      "step": 38460
    },
    {
      "epoch": 9.155686484556329,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1008,
      "step": 38461
    },
    {
      "epoch": 9.155924537285008,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 38462
    },
    {
      "epoch": 9.156162590013688,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8315,
      "step": 38463
    },
    {
      "epoch": 9.156400642742367,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1264,
      "step": 38464
    },
    {
      "epoch": 9.156638695471047,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0093,
      "step": 38465
    },
    {
      "epoch": 9.156876748199727,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0007,
      "step": 38466
    },
    {
      "epoch": 9.157114800928406,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1621,
      "step": 38467
    },
    {
      "epoch": 9.157352853657086,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9838,
      "step": 38468
    },
    {
      "epoch": 9.157590906385764,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9099,
      "step": 38469
    },
    {
      "epoch": 9.157828959114443,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 38470
    },
    {
      "epoch": 9.158067011843123,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 38471
    },
    {
      "epoch": 9.158305064571802,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 38472
    },
    {
      "epoch": 9.158543117300482,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 38473
    },
    {
      "epoch": 9.158781170029162,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 38474
    },
    {
      "epoch": 9.159019222757841,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 38475
    },
    {
      "epoch": 9.159257275486521,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7787,
      "step": 38476
    },
    {
      "epoch": 9.1594953282152,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 38477
    },
    {
      "epoch": 9.159733380943878,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 38478
    },
    {
      "epoch": 9.159971433672558,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8698,
      "step": 38479
    },
    {
      "epoch": 9.160209486401238,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9687,
      "step": 38480
    },
    {
      "epoch": 9.160447539129917,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7518,
      "step": 38481
    },
    {
      "epoch": 9.160685591858597,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1091,
      "step": 38482
    },
    {
      "epoch": 9.160923644587276,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 38483
    },
    {
      "epoch": 9.161161697315956,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.646,
      "step": 38484
    },
    {
      "epoch": 9.161399750044636,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 38485
    },
    {
      "epoch": 9.161637802773313,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1559,
      "step": 38486
    },
    {
      "epoch": 9.161875855501993,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0262,
      "step": 38487
    },
    {
      "epoch": 9.162113908230673,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2035,
      "step": 38488
    },
    {
      "epoch": 9.162351960959352,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 38489
    },
    {
      "epoch": 9.162590013688032,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1689,
      "step": 38490
    },
    {
      "epoch": 9.162828066416711,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 38491
    },
    {
      "epoch": 9.163066119145391,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1496,
      "step": 38492
    },
    {
      "epoch": 9.16330417187407,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 38493
    },
    {
      "epoch": 9.16354222460275,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 38494
    },
    {
      "epoch": 9.163780277331428,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 38495
    },
    {
      "epoch": 9.164018330060108,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 38496
    },
    {
      "epoch": 9.164256382788787,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.009,
      "step": 38497
    },
    {
      "epoch": 9.164494435517467,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 38498
    },
    {
      "epoch": 9.164732488246147,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 38499
    },
    {
      "epoch": 9.164970540974826,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 38500
    },
    {
      "epoch": 9.164970540974826,
      "eval_runtime": 0.5226,
      "eval_samples_per_second": 1.914,
      "eval_steps_per_second": 1.914,
      "step": 38500
    },
    {
      "epoch": 9.165208593703506,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1527,
      "step": 38501
    },
    {
      "epoch": 9.165446646432185,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9092,
      "step": 38502
    },
    {
      "epoch": 9.165684699160865,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0495,
      "step": 38503
    },
    {
      "epoch": 9.165922751889543,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.204,
      "step": 38504
    },
    {
      "epoch": 9.166160804618222,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 38505
    },
    {
      "epoch": 9.166398857346902,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9322,
      "step": 38506
    },
    {
      "epoch": 9.166636910075582,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0523,
      "step": 38507
    },
    {
      "epoch": 9.166874962804261,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.126,
      "step": 38508
    },
    {
      "epoch": 9.16711301553294,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0356,
      "step": 38509
    },
    {
      "epoch": 9.16735106826162,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.8071,
      "step": 38510
    },
    {
      "epoch": 9.1675891209903,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 38511
    },
    {
      "epoch": 9.16782717371898,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.0203,
      "step": 38512
    },
    {
      "epoch": 9.168065226447657,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.6564,
      "step": 38513
    },
    {
      "epoch": 9.168303279176337,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.123,
      "step": 38514
    },
    {
      "epoch": 9.168541331905017,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 38515
    },
    {
      "epoch": 9.168779384633696,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 2.2232,
      "step": 38516
    },
    {
      "epoch": 9.169017437362376,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 38517
    },
    {
      "epoch": 9.169255490091055,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 38518
    },
    {
      "epoch": 9.169493542819735,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.066,
      "step": 38519
    },
    {
      "epoch": 9.169731595548415,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.101,
      "step": 38520
    },
    {
      "epoch": 9.169969648277092,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1101,
      "step": 38521
    },
    {
      "epoch": 9.170207701005772,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7539,
      "step": 38522
    },
    {
      "epoch": 9.170445753734452,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7475,
      "step": 38523
    },
    {
      "epoch": 9.170683806463131,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0434,
      "step": 38524
    },
    {
      "epoch": 9.170921859191811,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 38525
    },
    {
      "epoch": 9.17115991192049,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 38526
    },
    {
      "epoch": 9.17139796464917,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 38527
    },
    {
      "epoch": 9.17163601737785,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0119,
      "step": 38528
    },
    {
      "epoch": 9.17187407010653,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 38529
    },
    {
      "epoch": 9.172112122835207,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2426,
      "step": 38530
    },
    {
      "epoch": 9.172350175563887,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1161,
      "step": 38531
    },
    {
      "epoch": 9.172588228292566,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0833,
      "step": 38532
    },
    {
      "epoch": 9.172826281021246,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 38533
    },
    {
      "epoch": 9.173064333749926,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2324,
      "step": 38534
    },
    {
      "epoch": 9.173302386478605,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0694,
      "step": 38535
    },
    {
      "epoch": 9.173540439207285,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2153,
      "step": 38536
    },
    {
      "epoch": 9.173778491935964,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 38537
    },
    {
      "epoch": 9.174016544664644,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 38538
    },
    {
      "epoch": 9.174254597393322,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8708,
      "step": 38539
    },
    {
      "epoch": 9.174492650122001,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 38540
    },
    {
      "epoch": 9.174730702850681,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 38541
    },
    {
      "epoch": 9.17496875557936,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 38542
    },
    {
      "epoch": 9.17520680830804,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8757,
      "step": 38543
    },
    {
      "epoch": 9.17544486103672,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 38544
    },
    {
      "epoch": 9.1756829137654,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 38545
    },
    {
      "epoch": 9.175920966494079,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.055,
      "step": 38546
    },
    {
      "epoch": 9.176159019222759,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8951,
      "step": 38547
    },
    {
      "epoch": 9.176397071951437,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2306,
      "step": 38548
    },
    {
      "epoch": 9.176635124680116,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 38549
    },
    {
      "epoch": 9.176873177408796,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1475,
      "step": 38550
    },
    {
      "epoch": 9.177111230137475,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9261,
      "step": 38551
    },
    {
      "epoch": 9.177349282866155,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 38552
    },
    {
      "epoch": 9.177587335594835,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 38553
    },
    {
      "epoch": 9.177825388323514,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.358,
      "step": 38554
    },
    {
      "epoch": 9.178063441052194,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9101,
      "step": 38555
    },
    {
      "epoch": 9.178301493780872,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.6093,
      "step": 38556
    },
    {
      "epoch": 9.178539546509551,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 38557
    },
    {
      "epoch": 9.17877759923823,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9712,
      "step": 38558
    },
    {
      "epoch": 9.17901565196691,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1227,
      "step": 38559
    },
    {
      "epoch": 9.17925370469559,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.199,
      "step": 38560
    },
    {
      "epoch": 9.17949175742427,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.166,
      "step": 38561
    },
    {
      "epoch": 9.17972981015295,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 38562
    },
    {
      "epoch": 9.179967862881629,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 38563
    },
    {
      "epoch": 9.180205915610308,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 38564
    },
    {
      "epoch": 9.180443968338986,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9319,
      "step": 38565
    },
    {
      "epoch": 9.180682021067666,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 38566
    },
    {
      "epoch": 9.180920073796345,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9135,
      "step": 38567
    },
    {
      "epoch": 9.181158126525025,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 38568
    },
    {
      "epoch": 9.181396179253705,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0344,
      "step": 38569
    },
    {
      "epoch": 9.181634231982384,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9344,
      "step": 38570
    },
    {
      "epoch": 9.181872284711064,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 38571
    },
    {
      "epoch": 9.182110337439743,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 38572
    },
    {
      "epoch": 9.182348390168423,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 38573
    },
    {
      "epoch": 9.182586442897101,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9238,
      "step": 38574
    },
    {
      "epoch": 9.18282449562578,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 38575
    },
    {
      "epoch": 9.18306254835446,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 38576
    },
    {
      "epoch": 9.18330060108314,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 38577
    },
    {
      "epoch": 9.18353865381182,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 38578
    },
    {
      "epoch": 9.183776706540499,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2505,
      "step": 38579
    },
    {
      "epoch": 9.184014759269179,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 38580
    },
    {
      "epoch": 9.184252811997858,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 38581
    },
    {
      "epoch": 9.184490864726538,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2778,
      "step": 38582
    },
    {
      "epoch": 9.184728917455216,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9982,
      "step": 38583
    },
    {
      "epoch": 9.184966970183895,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 38584
    },
    {
      "epoch": 9.185205022912575,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2125,
      "step": 38585
    },
    {
      "epoch": 9.185443075641254,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 38586
    },
    {
      "epoch": 9.185681128369934,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0384,
      "step": 38587
    },
    {
      "epoch": 9.185919181098614,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1784,
      "step": 38588
    },
    {
      "epoch": 9.186157233827293,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 38589
    },
    {
      "epoch": 9.186395286555973,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.756,
      "step": 38590
    },
    {
      "epoch": 9.186633339284652,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 38591
    },
    {
      "epoch": 9.18687139201333,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 38592
    },
    {
      "epoch": 9.18710944474201,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 38593
    },
    {
      "epoch": 9.18734749747069,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9425,
      "step": 38594
    },
    {
      "epoch": 9.187585550199369,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1832,
      "step": 38595
    },
    {
      "epoch": 9.187823602928049,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 38596
    },
    {
      "epoch": 9.188061655656728,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1619,
      "step": 38597
    },
    {
      "epoch": 9.188299708385408,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.5721,
      "step": 38598
    },
    {
      "epoch": 9.188537761114087,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9056,
      "step": 38599
    },
    {
      "epoch": 9.188775813842765,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1959,
      "step": 38600
    },
    {
      "epoch": 9.188775813842765,
      "eval_runtime": 0.5228,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 1.913,
      "step": 38600
    },
    {
      "epoch": 9.189013866571445,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 38601
    },
    {
      "epoch": 9.189251919300125,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8737,
      "step": 38602
    },
    {
      "epoch": 9.189489972028804,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 38603
    },
    {
      "epoch": 9.189728024757484,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 38604
    },
    {
      "epoch": 9.189966077486163,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2417,
      "step": 38605
    },
    {
      "epoch": 9.190204130214843,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8587,
      "step": 38606
    },
    {
      "epoch": 9.190442182943523,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 38607
    },
    {
      "epoch": 9.190680235672202,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.188,
      "step": 38608
    },
    {
      "epoch": 9.19091828840088,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8719,
      "step": 38609
    },
    {
      "epoch": 9.19115634112956,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 38610
    },
    {
      "epoch": 9.19139439385824,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1815,
      "step": 38611
    },
    {
      "epoch": 9.191632446586919,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 38612
    },
    {
      "epoch": 9.191870499315598,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8116,
      "step": 38613
    },
    {
      "epoch": 9.192108552044278,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9783,
      "step": 38614
    },
    {
      "epoch": 9.192346604772958,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7348,
      "step": 38615
    },
    {
      "epoch": 9.192584657501637,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 38616
    },
    {
      "epoch": 9.192822710230317,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 38617
    },
    {
      "epoch": 9.193060762958995,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 38618
    },
    {
      "epoch": 9.193298815687674,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0114,
      "step": 38619
    },
    {
      "epoch": 9.193536868416354,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1352,
      "step": 38620
    },
    {
      "epoch": 9.193774921145033,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 38621
    },
    {
      "epoch": 9.194012973873713,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1558,
      "step": 38622
    },
    {
      "epoch": 9.194251026602393,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.319,
      "step": 38623
    },
    {
      "epoch": 9.194489079331072,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1173,
      "step": 38624
    },
    {
      "epoch": 9.194727132059752,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 38625
    },
    {
      "epoch": 9.194965184788432,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 38626
    },
    {
      "epoch": 9.19520323751711,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 38627
    },
    {
      "epoch": 9.195441290245789,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0171,
      "step": 38628
    },
    {
      "epoch": 9.195679342974469,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1528,
      "step": 38629
    },
    {
      "epoch": 9.195917395703148,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 38630
    },
    {
      "epoch": 9.196155448431828,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 38631
    },
    {
      "epoch": 9.196393501160507,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0011,
      "step": 38632
    },
    {
      "epoch": 9.196631553889187,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0792,
      "step": 38633
    },
    {
      "epoch": 9.196869606617867,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 38634
    },
    {
      "epoch": 9.197107659346544,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.345,
      "step": 38635
    },
    {
      "epoch": 9.197345712075224,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 38636
    },
    {
      "epoch": 9.197583764803904,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.8441,
      "step": 38637
    },
    {
      "epoch": 9.197821817532583,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9042,
      "step": 38638
    },
    {
      "epoch": 9.198059870261263,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4617,
      "step": 38639
    },
    {
      "epoch": 9.198297922989942,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1404,
      "step": 38640
    },
    {
      "epoch": 9.198535975718622,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.121,
      "step": 38641
    },
    {
      "epoch": 9.198774028447302,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.4432,
      "step": 38642
    },
    {
      "epoch": 9.199012081175981,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8257,
      "step": 38643
    },
    {
      "epoch": 9.199250133904659,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 38644
    },
    {
      "epoch": 9.199488186633339,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8874,
      "step": 38645
    },
    {
      "epoch": 9.199726239362018,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 38646
    },
    {
      "epoch": 9.199964292090698,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9843,
      "step": 38647
    },
    {
      "epoch": 9.200202344819377,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 38648
    },
    {
      "epoch": 9.200440397548057,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 38649
    },
    {
      "epoch": 9.200678450276737,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 38650
    },
    {
      "epoch": 9.200916503005416,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 38651
    },
    {
      "epoch": 9.201154555734096,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 38652
    },
    {
      "epoch": 9.201392608462774,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 38653
    },
    {
      "epoch": 9.201630661191453,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 38654
    },
    {
      "epoch": 9.201868713920133,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1608,
      "step": 38655
    },
    {
      "epoch": 9.202106766648813,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.4704,
      "step": 38656
    },
    {
      "epoch": 9.202344819377492,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2036,
      "step": 38657
    },
    {
      "epoch": 9.202582872106172,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 38658
    },
    {
      "epoch": 9.202820924834851,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2655,
      "step": 38659
    },
    {
      "epoch": 9.203058977563531,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 38660
    },
    {
      "epoch": 9.20329703029221,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3141,
      "step": 38661
    },
    {
      "epoch": 9.203535083020888,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7451,
      "step": 38662
    },
    {
      "epoch": 9.203773135749568,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1337,
      "step": 38663
    },
    {
      "epoch": 9.204011188478248,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1268,
      "step": 38664
    },
    {
      "epoch": 9.204249241206927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 38665
    },
    {
      "epoch": 9.204487293935607,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 38666
    },
    {
      "epoch": 9.204725346664286,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8912,
      "step": 38667
    },
    {
      "epoch": 9.204963399392966,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 38668
    },
    {
      "epoch": 9.205201452121646,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9124,
      "step": 38669
    },
    {
      "epoch": 9.205439504850323,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0953,
      "step": 38670
    },
    {
      "epoch": 9.205677557579003,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 38671
    },
    {
      "epoch": 9.205915610307683,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 38672
    },
    {
      "epoch": 9.206153663036362,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1571,
      "step": 38673
    },
    {
      "epoch": 9.206391715765042,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 38674
    },
    {
      "epoch": 9.206629768493721,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.3266,
      "step": 38675
    },
    {
      "epoch": 9.206867821222401,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9302,
      "step": 38676
    },
    {
      "epoch": 9.20710587395108,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 38677
    },
    {
      "epoch": 9.20734392667976,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 38678
    },
    {
      "epoch": 9.207581979408438,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0479,
      "step": 38679
    },
    {
      "epoch": 9.207820032137118,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8574,
      "step": 38680
    },
    {
      "epoch": 9.208058084865797,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 38681
    },
    {
      "epoch": 9.208296137594477,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 38682
    },
    {
      "epoch": 9.208534190323157,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8215,
      "step": 38683
    },
    {
      "epoch": 9.208772243051836,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 38684
    },
    {
      "epoch": 9.209010295780516,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 38685
    },
    {
      "epoch": 9.209248348509195,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9165,
      "step": 38686
    },
    {
      "epoch": 9.209486401237875,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1432,
      "step": 38687
    },
    {
      "epoch": 9.209724453966553,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 38688
    },
    {
      "epoch": 9.209962506695232,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1495,
      "step": 38689
    },
    {
      "epoch": 9.210200559423912,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 38690
    },
    {
      "epoch": 9.210438612152592,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 38691
    },
    {
      "epoch": 9.210676664881271,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0478,
      "step": 38692
    },
    {
      "epoch": 9.21091471760995,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0956,
      "step": 38693
    },
    {
      "epoch": 9.21115277033863,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.3625,
      "step": 38694
    },
    {
      "epoch": 9.21139082306731,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1038,
      "step": 38695
    },
    {
      "epoch": 9.21162887579599,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8778,
      "step": 38696
    },
    {
      "epoch": 9.211866928524667,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 38697
    },
    {
      "epoch": 9.212104981253347,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8313,
      "step": 38698
    },
    {
      "epoch": 9.212343033982027,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2126,
      "step": 38699
    },
    {
      "epoch": 9.212581086710706,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9784,
      "step": 38700
    },
    {
      "epoch": 9.212581086710706,
      "eval_runtime": 0.5149,
      "eval_samples_per_second": 1.942,
      "eval_steps_per_second": 1.942,
      "step": 38700
    },
    {
      "epoch": 9.212819139439386,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.201,
      "step": 38701
    },
    {
      "epoch": 9.213057192168066,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8596,
      "step": 38702
    },
    {
      "epoch": 9.213295244896745,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 38703
    },
    {
      "epoch": 9.213533297625425,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 38704
    },
    {
      "epoch": 9.213771350354104,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 38705
    },
    {
      "epoch": 9.214009403082782,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 38706
    },
    {
      "epoch": 9.214247455811462,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 38707
    },
    {
      "epoch": 9.214485508540141,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 38708
    },
    {
      "epoch": 9.214723561268821,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8417,
      "step": 38709
    },
    {
      "epoch": 9.2149616139975,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 38710
    },
    {
      "epoch": 9.21519966672618,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8069,
      "step": 38711
    },
    {
      "epoch": 9.21543771945486,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9567,
      "step": 38712
    },
    {
      "epoch": 9.21567577218354,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 38713
    },
    {
      "epoch": 9.215913824912217,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.894,
      "step": 38714
    },
    {
      "epoch": 9.216151877640897,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8334,
      "step": 38715
    },
    {
      "epoch": 9.216389930369576,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9423,
      "step": 38716
    },
    {
      "epoch": 9.216627983098256,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1554,
      "step": 38717
    },
    {
      "epoch": 9.216866035826936,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0096,
      "step": 38718
    },
    {
      "epoch": 9.217104088555615,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.976,
      "step": 38719
    },
    {
      "epoch": 9.217342141284295,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0492,
      "step": 38720
    },
    {
      "epoch": 9.217580194012974,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8138,
      "step": 38721
    },
    {
      "epoch": 9.217818246741654,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8633,
      "step": 38722
    },
    {
      "epoch": 9.218056299470332,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0974,
      "step": 38723
    },
    {
      "epoch": 9.218294352199011,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1408,
      "step": 38724
    },
    {
      "epoch": 9.218532404927691,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.798,
      "step": 38725
    },
    {
      "epoch": 9.21877045765637,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7497,
      "step": 38726
    },
    {
      "epoch": 9.21900851038505,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3355,
      "step": 38727
    },
    {
      "epoch": 9.21924656311373,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 38728
    },
    {
      "epoch": 9.21948461584241,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7567,
      "step": 38729
    },
    {
      "epoch": 9.21972266857109,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0827,
      "step": 38730
    },
    {
      "epoch": 9.219960721299769,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 38731
    },
    {
      "epoch": 9.220198774028447,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0656,
      "step": 38732
    },
    {
      "epoch": 9.220436826757126,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9599,
      "step": 38733
    },
    {
      "epoch": 9.220674879485806,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 38734
    },
    {
      "epoch": 9.220912932214485,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 38735
    },
    {
      "epoch": 9.221150984943165,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 38736
    },
    {
      "epoch": 9.221389037671845,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 38737
    },
    {
      "epoch": 9.221627090400524,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 38738
    },
    {
      "epoch": 9.221865143129204,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 38739
    },
    {
      "epoch": 9.222103195857883,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0408,
      "step": 38740
    },
    {
      "epoch": 9.222341248586561,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 38741
    },
    {
      "epoch": 9.22257930131524,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0028,
      "step": 38742
    },
    {
      "epoch": 9.22281735404392,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 38743
    },
    {
      "epoch": 9.2230554067726,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 38744
    },
    {
      "epoch": 9.22329345950128,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1844,
      "step": 38745
    },
    {
      "epoch": 9.22353151222996,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 38746
    },
    {
      "epoch": 9.223769564958639,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9729,
      "step": 38747
    },
    {
      "epoch": 9.224007617687318,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.928,
      "step": 38748
    },
    {
      "epoch": 9.224245670415996,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8614,
      "step": 38749
    },
    {
      "epoch": 9.224483723144676,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 38750
    },
    {
      "epoch": 9.224721775873356,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 38751
    },
    {
      "epoch": 9.224959828602035,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2853,
      "step": 38752
    },
    {
      "epoch": 9.225197881330715,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1551,
      "step": 38753
    },
    {
      "epoch": 9.225435934059394,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2603,
      "step": 38754
    },
    {
      "epoch": 9.225673986788074,
      "grad_norm": 2.875,
      "learning_rate": 3e-05,
      "loss": 1.9465,
      "step": 38755
    },
    {
      "epoch": 9.225912039516754,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.6291,
      "step": 38756
    },
    {
      "epoch": 9.226150092245433,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.987,
      "step": 38757
    },
    {
      "epoch": 9.226388144974111,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7484,
      "step": 38758
    },
    {
      "epoch": 9.22662619770279,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 38759
    },
    {
      "epoch": 9.22686425043147,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.13,
      "step": 38760
    },
    {
      "epoch": 9.22710230316015,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2419,
      "step": 38761
    },
    {
      "epoch": 9.22734035588883,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1402,
      "step": 38762
    },
    {
      "epoch": 9.227578408617509,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 38763
    },
    {
      "epoch": 9.227816461346189,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7476,
      "step": 38764
    },
    {
      "epoch": 9.228054514074868,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9093,
      "step": 38765
    },
    {
      "epoch": 9.228292566803548,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 1.7847,
      "step": 38766
    },
    {
      "epoch": 9.228530619532226,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1862,
      "step": 38767
    },
    {
      "epoch": 9.228768672260905,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 38768
    },
    {
      "epoch": 9.229006724989585,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0878,
      "step": 38769
    },
    {
      "epoch": 9.229244777718264,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 38770
    },
    {
      "epoch": 9.229482830446944,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 38771
    },
    {
      "epoch": 9.229720883175624,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0634,
      "step": 38772
    },
    {
      "epoch": 9.229958935904303,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9632,
      "step": 38773
    },
    {
      "epoch": 9.230196988632983,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 38774
    },
    {
      "epoch": 9.230435041361662,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 38775
    },
    {
      "epoch": 9.23067309409034,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 38776
    },
    {
      "epoch": 9.23091114681902,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8452,
      "step": 38777
    },
    {
      "epoch": 9.2311491995477,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8249,
      "step": 38778
    },
    {
      "epoch": 9.23138725227638,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.799,
      "step": 38779
    },
    {
      "epoch": 9.231625305005059,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.4114,
      "step": 38780
    },
    {
      "epoch": 9.231863357733738,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8702,
      "step": 38781
    },
    {
      "epoch": 9.232101410462418,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 38782
    },
    {
      "epoch": 9.232339463191098,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 38783
    },
    {
      "epoch": 9.232577515919775,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9362,
      "step": 38784
    },
    {
      "epoch": 9.232815568648455,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8831,
      "step": 38785
    },
    {
      "epoch": 9.233053621377135,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9679,
      "step": 38786
    },
    {
      "epoch": 9.233291674105814,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0035,
      "step": 38787
    },
    {
      "epoch": 9.233529726834494,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2476,
      "step": 38788
    },
    {
      "epoch": 9.233767779563173,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 38789
    },
    {
      "epoch": 9.234005832291853,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.884,
      "step": 38790
    },
    {
      "epoch": 9.234243885020533,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1789,
      "step": 38791
    },
    {
      "epoch": 9.234481937749212,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 38792
    },
    {
      "epoch": 9.23471999047789,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 38793
    },
    {
      "epoch": 9.23495804320657,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.693,
      "step": 38794
    },
    {
      "epoch": 9.23519609593525,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3293,
      "step": 38795
    },
    {
      "epoch": 9.235434148663929,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 38796
    },
    {
      "epoch": 9.235672201392608,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8608,
      "step": 38797
    },
    {
      "epoch": 9.235910254121288,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9281,
      "step": 38798
    },
    {
      "epoch": 9.236148306849968,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8895,
      "step": 38799
    },
    {
      "epoch": 9.236386359578647,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 38800
    },
    {
      "epoch": 9.236386359578647,
      "eval_runtime": 0.5186,
      "eval_samples_per_second": 1.928,
      "eval_steps_per_second": 1.928,
      "step": 38800
    },
    {
      "epoch": 9.236624412307327,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 38801
    },
    {
      "epoch": 9.236862465036005,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 38802
    },
    {
      "epoch": 9.237100517764684,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0906,
      "step": 38803
    },
    {
      "epoch": 9.237338570493364,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0643,
      "step": 38804
    },
    {
      "epoch": 9.237576623222044,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9039,
      "step": 38805
    },
    {
      "epoch": 9.237814675950723,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9584,
      "step": 38806
    },
    {
      "epoch": 9.238052728679403,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 38807
    },
    {
      "epoch": 9.238290781408082,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3742,
      "step": 38808
    },
    {
      "epoch": 9.238528834136762,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0029,
      "step": 38809
    },
    {
      "epoch": 9.238766886865442,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8673,
      "step": 38810
    },
    {
      "epoch": 9.23900493959412,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.2481,
      "step": 38811
    },
    {
      "epoch": 9.239242992322799,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9936,
      "step": 38812
    },
    {
      "epoch": 9.239481045051479,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1168,
      "step": 38813
    },
    {
      "epoch": 9.239719097780158,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 38814
    },
    {
      "epoch": 9.239957150508838,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9424,
      "step": 38815
    },
    {
      "epoch": 9.240195203237517,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2386,
      "step": 38816
    },
    {
      "epoch": 9.240433255966197,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7805,
      "step": 38817
    },
    {
      "epoch": 9.240671308694877,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0852,
      "step": 38818
    },
    {
      "epoch": 9.240909361423554,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.01,
      "step": 38819
    },
    {
      "epoch": 9.241147414152234,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 38820
    },
    {
      "epoch": 9.241385466880914,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 38821
    },
    {
      "epoch": 9.241623519609593,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0514,
      "step": 38822
    },
    {
      "epoch": 9.241861572338273,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8805,
      "step": 38823
    },
    {
      "epoch": 9.242099625066952,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 38824
    },
    {
      "epoch": 9.242337677795632,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2464,
      "step": 38825
    },
    {
      "epoch": 9.242575730524312,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1785,
      "step": 38826
    },
    {
      "epoch": 9.242813783252991,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 38827
    },
    {
      "epoch": 9.24305183598167,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 38828
    },
    {
      "epoch": 9.243289888710349,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1753,
      "step": 38829
    },
    {
      "epoch": 9.243527941439028,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.484,
      "step": 38830
    },
    {
      "epoch": 9.243765994167708,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.05,
      "step": 38831
    },
    {
      "epoch": 9.244004046896388,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0873,
      "step": 38832
    },
    {
      "epoch": 9.244242099625067,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0167,
      "step": 38833
    },
    {
      "epoch": 9.244480152353747,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9518,
      "step": 38834
    },
    {
      "epoch": 9.244718205082426,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.4082,
      "step": 38835
    },
    {
      "epoch": 9.244956257811106,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8653,
      "step": 38836
    },
    {
      "epoch": 9.245194310539784,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 38837
    },
    {
      "epoch": 9.245432363268463,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9516,
      "step": 38838
    },
    {
      "epoch": 9.245670415997143,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8513,
      "step": 38839
    },
    {
      "epoch": 9.245908468725823,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.3246,
      "step": 38840
    },
    {
      "epoch": 9.246146521454502,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.715,
      "step": 38841
    },
    {
      "epoch": 9.246384574183182,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 38842
    },
    {
      "epoch": 9.246622626911861,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8929,
      "step": 38843
    },
    {
      "epoch": 9.246860679640541,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.096,
      "step": 38844
    },
    {
      "epoch": 9.24709873236922,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 38845
    },
    {
      "epoch": 9.247336785097898,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8779,
      "step": 38846
    },
    {
      "epoch": 9.247574837826578,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.891,
      "step": 38847
    },
    {
      "epoch": 9.247812890555258,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2334,
      "step": 38848
    },
    {
      "epoch": 9.248050943283937,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.5811,
      "step": 38849
    },
    {
      "epoch": 9.248288996012617,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9878,
      "step": 38850
    },
    {
      "epoch": 9.248527048741296,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.922,
      "step": 38851
    },
    {
      "epoch": 9.248765101469976,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 38852
    },
    {
      "epoch": 9.249003154198656,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9204,
      "step": 38853
    },
    {
      "epoch": 9.249241206927334,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 38854
    },
    {
      "epoch": 9.249479259656013,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3243,
      "step": 38855
    },
    {
      "epoch": 9.249717312384693,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9217,
      "step": 38856
    },
    {
      "epoch": 9.249955365113372,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 38857
    },
    {
      "epoch": 9.250193417842052,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9464,
      "step": 38858
    },
    {
      "epoch": 9.250431470570732,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9591,
      "step": 38859
    },
    {
      "epoch": 9.250669523299411,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.1208,
      "step": 38860
    },
    {
      "epoch": 9.25090757602809,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 38861
    },
    {
      "epoch": 9.25114562875677,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3378,
      "step": 38862
    },
    {
      "epoch": 9.251383681485448,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9412,
      "step": 38863
    },
    {
      "epoch": 9.251621734214128,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9719,
      "step": 38864
    },
    {
      "epoch": 9.251859786942807,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 38865
    },
    {
      "epoch": 9.252097839671487,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0921,
      "step": 38866
    },
    {
      "epoch": 9.252335892400167,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 38867
    },
    {
      "epoch": 9.252573945128846,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8732,
      "step": 38868
    },
    {
      "epoch": 9.252811997857526,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0283,
      "step": 38869
    },
    {
      "epoch": 9.253050050586205,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 38870
    },
    {
      "epoch": 9.253288103314885,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 38871
    },
    {
      "epoch": 9.253526156043563,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 38872
    },
    {
      "epoch": 9.253764208772242,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8881,
      "step": 38873
    },
    {
      "epoch": 9.254002261500922,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0779,
      "step": 38874
    },
    {
      "epoch": 9.254240314229602,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8387,
      "step": 38875
    },
    {
      "epoch": 9.254478366958281,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.106,
      "step": 38876
    },
    {
      "epoch": 9.254716419686961,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0051,
      "step": 38877
    },
    {
      "epoch": 9.25495447241564,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1145,
      "step": 38878
    },
    {
      "epoch": 9.25519252514432,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1016,
      "step": 38879
    },
    {
      "epoch": 9.255430577873,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.179,
      "step": 38880
    },
    {
      "epoch": 9.255668630601678,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0602,
      "step": 38881
    },
    {
      "epoch": 9.255906683330357,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9347,
      "step": 38882
    },
    {
      "epoch": 9.256144736059037,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9951,
      "step": 38883
    },
    {
      "epoch": 9.256382788787716,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 38884
    },
    {
      "epoch": 9.256620841516396,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0964,
      "step": 38885
    },
    {
      "epoch": 9.256858894245076,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9609,
      "step": 38886
    },
    {
      "epoch": 9.257096946973755,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.2385,
      "step": 38887
    },
    {
      "epoch": 9.257334999702435,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9983,
      "step": 38888
    },
    {
      "epoch": 9.257573052431113,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1312,
      "step": 38889
    },
    {
      "epoch": 9.257811105159792,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 38890
    },
    {
      "epoch": 9.258049157888472,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2207,
      "step": 38891
    },
    {
      "epoch": 9.258287210617151,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0984,
      "step": 38892
    },
    {
      "epoch": 9.258525263345831,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 38893
    },
    {
      "epoch": 9.25876331607451,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.082,
      "step": 38894
    },
    {
      "epoch": 9.25900136880319,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3051,
      "step": 38895
    },
    {
      "epoch": 9.25923942153187,
      "grad_norm": 6.28125,
      "learning_rate": 3e-05,
      "loss": 2.0576,
      "step": 38896
    },
    {
      "epoch": 9.25947747426055,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.029,
      "step": 38897
    },
    {
      "epoch": 9.259715526989227,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1995,
      "step": 38898
    },
    {
      "epoch": 9.259953579717907,
      "grad_norm": 6.5,
      "learning_rate": 3e-05,
      "loss": 1.821,
      "step": 38899
    },
    {
      "epoch": 9.260191632446586,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1102,
      "step": 38900
    },
    {
      "epoch": 9.260191632446586,
      "eval_runtime": 0.5122,
      "eval_samples_per_second": 1.952,
      "eval_steps_per_second": 1.952,
      "step": 38900
    },
    {
      "epoch": 9.260429685175266,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 38901
    },
    {
      "epoch": 9.260667737903946,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9508,
      "step": 38902
    },
    {
      "epoch": 9.260905790632625,
      "grad_norm": 6.09375,
      "learning_rate": 3e-05,
      "loss": 2.0667,
      "step": 38903
    },
    {
      "epoch": 9.261143843361305,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 38904
    },
    {
      "epoch": 9.261381896089985,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 38905
    },
    {
      "epoch": 9.261619948818664,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 2.1458,
      "step": 38906
    },
    {
      "epoch": 9.261858001547342,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0464,
      "step": 38907
    },
    {
      "epoch": 9.262096054276022,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 38908
    },
    {
      "epoch": 9.262334107004701,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1133,
      "step": 38909
    },
    {
      "epoch": 9.26257215973338,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.0589,
      "step": 38910
    },
    {
      "epoch": 9.26281021246206,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 38911
    },
    {
      "epoch": 9.26304826519074,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 38912
    },
    {
      "epoch": 9.26328631791942,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0766,
      "step": 38913
    },
    {
      "epoch": 9.2635243706481,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 38914
    },
    {
      "epoch": 9.263762423376779,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0213,
      "step": 38915
    },
    {
      "epoch": 9.264000476105457,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0322,
      "step": 38916
    },
    {
      "epoch": 9.264238528834136,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1435,
      "step": 38917
    },
    {
      "epoch": 9.264476581562816,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 38918
    },
    {
      "epoch": 9.264714634291495,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 38919
    },
    {
      "epoch": 9.264952687020175,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 38920
    },
    {
      "epoch": 9.265190739748855,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 38921
    },
    {
      "epoch": 9.265428792477534,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 38922
    },
    {
      "epoch": 9.265666845206214,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7743,
      "step": 38923
    },
    {
      "epoch": 9.265904897934892,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 38924
    },
    {
      "epoch": 9.266142950663571,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8666,
      "step": 38925
    },
    {
      "epoch": 9.266381003392251,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.4037,
      "step": 38926
    },
    {
      "epoch": 9.26661905612093,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0161,
      "step": 38927
    },
    {
      "epoch": 9.26685710884961,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.7495,
      "step": 38928
    },
    {
      "epoch": 9.26709516157829,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 38929
    },
    {
      "epoch": 9.26733321430697,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 38930
    },
    {
      "epoch": 9.267571267035649,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 38931
    },
    {
      "epoch": 9.267809319764329,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 38932
    },
    {
      "epoch": 9.268047372493006,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8933,
      "step": 38933
    },
    {
      "epoch": 9.268285425221686,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8075,
      "step": 38934
    },
    {
      "epoch": 9.268523477950366,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8866,
      "step": 38935
    },
    {
      "epoch": 9.268761530679045,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1349,
      "step": 38936
    },
    {
      "epoch": 9.268999583407725,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 38937
    },
    {
      "epoch": 9.269237636136404,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.2462,
      "step": 38938
    },
    {
      "epoch": 9.269475688865084,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 38939
    },
    {
      "epoch": 9.269713741593764,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 38940
    },
    {
      "epoch": 9.269951794322443,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1018,
      "step": 38941
    },
    {
      "epoch": 9.270189847051121,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 38942
    },
    {
      "epoch": 9.2704278997798,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8103,
      "step": 38943
    },
    {
      "epoch": 9.27066595250848,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0394,
      "step": 38944
    },
    {
      "epoch": 9.27090400523716,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1806,
      "step": 38945
    },
    {
      "epoch": 9.27114205796584,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.8589,
      "step": 38946
    },
    {
      "epoch": 9.271380110694519,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9012,
      "step": 38947
    },
    {
      "epoch": 9.271618163423199,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1185,
      "step": 38948
    },
    {
      "epoch": 9.271856216151878,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9596,
      "step": 38949
    },
    {
      "epoch": 9.272094268880558,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1154,
      "step": 38950
    },
    {
      "epoch": 9.272332321609236,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9398,
      "step": 38951
    },
    {
      "epoch": 9.272570374337915,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9284,
      "step": 38952
    },
    {
      "epoch": 9.272808427066595,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0244,
      "step": 38953
    },
    {
      "epoch": 9.273046479795275,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.851,
      "step": 38954
    },
    {
      "epoch": 9.273284532523954,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.7712,
      "step": 38955
    },
    {
      "epoch": 9.273522585252634,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.7271,
      "step": 38956
    },
    {
      "epoch": 9.273760637981313,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7349,
      "step": 38957
    },
    {
      "epoch": 9.273998690709993,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.4831,
      "step": 38958
    },
    {
      "epoch": 9.27423674343867,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 38959
    },
    {
      "epoch": 9.27447479616735,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 38960
    },
    {
      "epoch": 9.27471284889603,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4103,
      "step": 38961
    },
    {
      "epoch": 9.27495090162471,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8715,
      "step": 38962
    },
    {
      "epoch": 9.27518895435339,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9876,
      "step": 38963
    },
    {
      "epoch": 9.275427007082069,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.2014,
      "step": 38964
    },
    {
      "epoch": 9.275665059810748,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9346,
      "step": 38965
    },
    {
      "epoch": 9.275903112539428,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2546,
      "step": 38966
    },
    {
      "epoch": 9.276141165268108,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 38967
    },
    {
      "epoch": 9.276379217996785,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0313,
      "step": 38968
    },
    {
      "epoch": 9.276617270725465,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.8309,
      "step": 38969
    },
    {
      "epoch": 9.276855323454145,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9083,
      "step": 38970
    },
    {
      "epoch": 9.277093376182824,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0438,
      "step": 38971
    },
    {
      "epoch": 9.277331428911504,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 38972
    },
    {
      "epoch": 9.277569481640183,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 38973
    },
    {
      "epoch": 9.277807534368863,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 38974
    },
    {
      "epoch": 9.278045587097543,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 38975
    },
    {
      "epoch": 9.278283639826222,
      "grad_norm": 2.8125,
      "learning_rate": 3e-05,
      "loss": 1.9905,
      "step": 38976
    },
    {
      "epoch": 9.2785216925549,
      "grad_norm": 6.0625,
      "learning_rate": 3e-05,
      "loss": 2.1591,
      "step": 38977
    },
    {
      "epoch": 9.27875974528358,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.9134,
      "step": 38978
    },
    {
      "epoch": 9.27899779801226,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 38979
    },
    {
      "epoch": 9.279235850740939,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 38980
    },
    {
      "epoch": 9.279473903469619,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.046,
      "step": 38981
    },
    {
      "epoch": 9.279711956198298,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.7777,
      "step": 38982
    },
    {
      "epoch": 9.279950008926978,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.214,
      "step": 38983
    },
    {
      "epoch": 9.280188061655657,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 38984
    },
    {
      "epoch": 9.280426114384337,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 38985
    },
    {
      "epoch": 9.280664167113015,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1501,
      "step": 38986
    },
    {
      "epoch": 9.280902219841694,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0079,
      "step": 38987
    },
    {
      "epoch": 9.281140272570374,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 38988
    },
    {
      "epoch": 9.281378325299054,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2539,
      "step": 38989
    },
    {
      "epoch": 9.281616378027733,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 38990
    },
    {
      "epoch": 9.281854430756413,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 38991
    },
    {
      "epoch": 9.282092483485092,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9413,
      "step": 38992
    },
    {
      "epoch": 9.282330536213772,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8905,
      "step": 38993
    },
    {
      "epoch": 9.28256858894245,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 38994
    },
    {
      "epoch": 9.28280664167113,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 38995
    },
    {
      "epoch": 9.283044694399809,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9672,
      "step": 38996
    },
    {
      "epoch": 9.283282747128489,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 38997
    },
    {
      "epoch": 9.283520799857168,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1624,
      "step": 38998
    },
    {
      "epoch": 9.283758852585848,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0699,
      "step": 38999
    },
    {
      "epoch": 9.283996905314527,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 39000
    },
    {
      "epoch": 9.283996905314527,
      "eval_runtime": 0.5175,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 39000
    },
    {
      "epoch": 9.284234958043207,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0879,
      "step": 39001
    },
    {
      "epoch": 9.284473010771887,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 39002
    },
    {
      "epoch": 9.284711063500565,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9494,
      "step": 39003
    },
    {
      "epoch": 9.284949116229244,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 39004
    },
    {
      "epoch": 9.285187168957924,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0746,
      "step": 39005
    },
    {
      "epoch": 9.285425221686603,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0584,
      "step": 39006
    },
    {
      "epoch": 9.285663274415283,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0644,
      "step": 39007
    },
    {
      "epoch": 9.285901327143963,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 39008
    },
    {
      "epoch": 9.286139379872642,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 39009
    },
    {
      "epoch": 9.286377432601322,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0024,
      "step": 39010
    },
    {
      "epoch": 9.286615485330001,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 39011
    },
    {
      "epoch": 9.28685353805868,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 39012
    },
    {
      "epoch": 9.287091590787359,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 39013
    },
    {
      "epoch": 9.287329643516038,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.4399,
      "step": 39014
    },
    {
      "epoch": 9.287567696244718,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.9688,
      "step": 39015
    },
    {
      "epoch": 9.287805748973398,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2425,
      "step": 39016
    },
    {
      "epoch": 9.288043801702077,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2716,
      "step": 39017
    },
    {
      "epoch": 9.288281854430757,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9334,
      "step": 39018
    },
    {
      "epoch": 9.288519907159436,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 39019
    },
    {
      "epoch": 9.288757959888116,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 39020
    },
    {
      "epoch": 9.288996012616794,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 39021
    },
    {
      "epoch": 9.289234065345473,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 39022
    },
    {
      "epoch": 9.289472118074153,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9912,
      "step": 39023
    },
    {
      "epoch": 9.289710170802833,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 39024
    },
    {
      "epoch": 9.289948223531512,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 39025
    },
    {
      "epoch": 9.290186276260192,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0264,
      "step": 39026
    },
    {
      "epoch": 9.290424328988871,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 39027
    },
    {
      "epoch": 9.290662381717551,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0848,
      "step": 39028
    },
    {
      "epoch": 9.290900434446229,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2765,
      "step": 39029
    },
    {
      "epoch": 9.291138487174909,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.1048,
      "step": 39030
    },
    {
      "epoch": 9.291376539903588,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0937,
      "step": 39031
    },
    {
      "epoch": 9.291614592632268,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8636,
      "step": 39032
    },
    {
      "epoch": 9.291852645360947,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0148,
      "step": 39033
    },
    {
      "epoch": 9.292090698089627,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.2056,
      "step": 39034
    },
    {
      "epoch": 9.292328750818307,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 39035
    },
    {
      "epoch": 9.292566803546986,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 39036
    },
    {
      "epoch": 9.292804856275666,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 2.1395,
      "step": 39037
    },
    {
      "epoch": 9.293042909004345,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.8109,
      "step": 39038
    },
    {
      "epoch": 9.293280961733023,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 39039
    },
    {
      "epoch": 9.293519014461703,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9145,
      "step": 39040
    },
    {
      "epoch": 9.293757067190382,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 39041
    },
    {
      "epoch": 9.293995119919062,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9818,
      "step": 39042
    },
    {
      "epoch": 9.294233172647742,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9911,
      "step": 39043
    },
    {
      "epoch": 9.294471225376421,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 39044
    },
    {
      "epoch": 9.2947092781051,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 39045
    },
    {
      "epoch": 9.29494733083378,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1562,
      "step": 39046
    },
    {
      "epoch": 9.295185383562458,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 39047
    },
    {
      "epoch": 9.295423436291138,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 39048
    },
    {
      "epoch": 9.295661489019817,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.274,
      "step": 39049
    },
    {
      "epoch": 9.295899541748497,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2879,
      "step": 39050
    },
    {
      "epoch": 9.296137594477177,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9253,
      "step": 39051
    },
    {
      "epoch": 9.296375647205856,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0082,
      "step": 39052
    },
    {
      "epoch": 9.296613699934536,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 39053
    },
    {
      "epoch": 9.296851752663216,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1323,
      "step": 39054
    },
    {
      "epoch": 9.297089805391895,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 39055
    },
    {
      "epoch": 9.297327858120573,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 39056
    },
    {
      "epoch": 9.297565910849253,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8168,
      "step": 39057
    },
    {
      "epoch": 9.297803963577932,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1052,
      "step": 39058
    },
    {
      "epoch": 9.298042016306612,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 39059
    },
    {
      "epoch": 9.298280069035291,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8882,
      "step": 39060
    },
    {
      "epoch": 9.298518121763971,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 39061
    },
    {
      "epoch": 9.29875617449265,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 39062
    },
    {
      "epoch": 9.29899422722133,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2668,
      "step": 39063
    },
    {
      "epoch": 9.29923227995001,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0294,
      "step": 39064
    },
    {
      "epoch": 9.299470332678688,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 39065
    },
    {
      "epoch": 9.299708385407367,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 39066
    },
    {
      "epoch": 9.299946438136047,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2149,
      "step": 39067
    },
    {
      "epoch": 9.300184490864726,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0339,
      "step": 39068
    },
    {
      "epoch": 9.300422543593406,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1744,
      "step": 39069
    },
    {
      "epoch": 9.300660596322086,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1598,
      "step": 39070
    },
    {
      "epoch": 9.300898649050765,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.9439,
      "step": 39071
    },
    {
      "epoch": 9.301136701779445,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8263,
      "step": 39072
    },
    {
      "epoch": 9.301374754508124,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2353,
      "step": 39073
    },
    {
      "epoch": 9.301612807236802,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 39074
    },
    {
      "epoch": 9.301850859965482,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0142,
      "step": 39075
    },
    {
      "epoch": 9.302088912694161,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9479,
      "step": 39076
    },
    {
      "epoch": 9.302326965422841,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.226,
      "step": 39077
    },
    {
      "epoch": 9.30256501815152,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.343,
      "step": 39078
    },
    {
      "epoch": 9.3028030708802,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 39079
    },
    {
      "epoch": 9.30304112360888,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 39080
    },
    {
      "epoch": 9.30327917633756,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0381,
      "step": 39081
    },
    {
      "epoch": 9.303517229066237,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 39082
    },
    {
      "epoch": 9.303755281794917,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8932,
      "step": 39083
    },
    {
      "epoch": 9.303993334523597,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 39084
    },
    {
      "epoch": 9.304231387252276,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0,
      "step": 39085
    },
    {
      "epoch": 9.304469439980956,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 39086
    },
    {
      "epoch": 9.304707492709635,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 39087
    },
    {
      "epoch": 9.304945545438315,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9901,
      "step": 39088
    },
    {
      "epoch": 9.305183598166995,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9127,
      "step": 39089
    },
    {
      "epoch": 9.305421650895674,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1045,
      "step": 39090
    },
    {
      "epoch": 9.305659703624352,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2369,
      "step": 39091
    },
    {
      "epoch": 9.305897756353032,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0617,
      "step": 39092
    },
    {
      "epoch": 9.306135809081711,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0724,
      "step": 39093
    },
    {
      "epoch": 9.30637386181039,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 39094
    },
    {
      "epoch": 9.30661191453907,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 39095
    },
    {
      "epoch": 9.30684996726775,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1661,
      "step": 39096
    },
    {
      "epoch": 9.30708801999643,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 39097
    },
    {
      "epoch": 9.30732607272511,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1651,
      "step": 39098
    },
    {
      "epoch": 9.307564125453789,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 39099
    },
    {
      "epoch": 9.307802178182467,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8455,
      "step": 39100
    },
    {
      "epoch": 9.307802178182467,
      "eval_runtime": 0.5243,
      "eval_samples_per_second": 1.907,
      "eval_steps_per_second": 1.907,
      "step": 39100
    },
    {
      "epoch": 9.308040230911146,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0341,
      "step": 39101
    },
    {
      "epoch": 9.308278283639826,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 39102
    },
    {
      "epoch": 9.308516336368506,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1035,
      "step": 39103
    },
    {
      "epoch": 9.308754389097185,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8319,
      "step": 39104
    },
    {
      "epoch": 9.308992441825865,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7995,
      "step": 39105
    },
    {
      "epoch": 9.309230494554544,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0991,
      "step": 39106
    },
    {
      "epoch": 9.309468547283224,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0853,
      "step": 39107
    },
    {
      "epoch": 9.309706600011904,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.8989,
      "step": 39108
    },
    {
      "epoch": 9.309944652740581,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0081,
      "step": 39109
    },
    {
      "epoch": 9.310182705469261,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9232,
      "step": 39110
    },
    {
      "epoch": 9.31042075819794,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 39111
    },
    {
      "epoch": 9.31065881092662,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0726,
      "step": 39112
    },
    {
      "epoch": 9.3108968636553,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8788,
      "step": 39113
    },
    {
      "epoch": 9.31113491638398,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2347,
      "step": 39114
    },
    {
      "epoch": 9.311372969112659,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 39115
    },
    {
      "epoch": 9.311611021841339,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 39116
    },
    {
      "epoch": 9.311849074570016,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8208,
      "step": 39117
    },
    {
      "epoch": 9.312087127298696,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1087,
      "step": 39118
    },
    {
      "epoch": 9.312325180027376,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6675,
      "step": 39119
    },
    {
      "epoch": 9.312563232756055,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9114,
      "step": 39120
    },
    {
      "epoch": 9.312801285484735,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 39121
    },
    {
      "epoch": 9.313039338213414,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0248,
      "step": 39122
    },
    {
      "epoch": 9.313277390942094,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9935,
      "step": 39123
    },
    {
      "epoch": 9.313515443670774,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9806,
      "step": 39124
    },
    {
      "epoch": 9.313753496399453,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.3034,
      "step": 39125
    },
    {
      "epoch": 9.313991549128131,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 39126
    },
    {
      "epoch": 9.31422960185681,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.7956,
      "step": 39127
    },
    {
      "epoch": 9.31446765458549,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.265,
      "step": 39128
    },
    {
      "epoch": 9.31470570731417,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0188,
      "step": 39129
    },
    {
      "epoch": 9.31494376004285,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 39130
    },
    {
      "epoch": 9.31518181277153,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8599,
      "step": 39131
    },
    {
      "epoch": 9.315419865500209,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 39132
    },
    {
      "epoch": 9.315657918228888,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1575,
      "step": 39133
    },
    {
      "epoch": 9.315895970957568,
      "grad_norm": 2.859375,
      "learning_rate": 3e-05,
      "loss": 1.8231,
      "step": 39134
    },
    {
      "epoch": 9.316134023686246,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3581,
      "step": 39135
    },
    {
      "epoch": 9.316372076414925,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.952,
      "step": 39136
    },
    {
      "epoch": 9.316610129143605,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 39137
    },
    {
      "epoch": 9.316848181872285,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 39138
    },
    {
      "epoch": 9.317086234600964,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 39139
    },
    {
      "epoch": 9.317324287329644,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 39140
    },
    {
      "epoch": 9.317562340058323,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.403,
      "step": 39141
    },
    {
      "epoch": 9.317800392787003,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0624,
      "step": 39142
    },
    {
      "epoch": 9.318038445515683,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 39143
    },
    {
      "epoch": 9.31827649824436,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1967,
      "step": 39144
    },
    {
      "epoch": 9.31851455097304,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2533,
      "step": 39145
    },
    {
      "epoch": 9.31875260370172,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1944,
      "step": 39146
    },
    {
      "epoch": 9.3189906564304,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0462,
      "step": 39147
    },
    {
      "epoch": 9.319228709159079,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 39148
    },
    {
      "epoch": 9.319466761887758,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2146,
      "step": 39149
    },
    {
      "epoch": 9.319704814616438,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.3194,
      "step": 39150
    },
    {
      "epoch": 9.319942867345118,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2433,
      "step": 39151
    },
    {
      "epoch": 9.320180920073796,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8368,
      "step": 39152
    },
    {
      "epoch": 9.320418972802475,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.5041,
      "step": 39153
    },
    {
      "epoch": 9.320657025531155,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 39154
    },
    {
      "epoch": 9.320895078259834,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2317,
      "step": 39155
    },
    {
      "epoch": 9.321133130988514,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8469,
      "step": 39156
    },
    {
      "epoch": 9.321371183717194,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.124,
      "step": 39157
    },
    {
      "epoch": 9.321609236445873,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9753,
      "step": 39158
    },
    {
      "epoch": 9.321847289174553,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 39159
    },
    {
      "epoch": 9.322085341903232,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 39160
    },
    {
      "epoch": 9.32232339463191,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 39161
    },
    {
      "epoch": 9.32256144736059,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0366,
      "step": 39162
    },
    {
      "epoch": 9.32279950008927,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 39163
    },
    {
      "epoch": 9.323037552817949,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.5154,
      "step": 39164
    },
    {
      "epoch": 9.323275605546629,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.7028,
      "step": 39165
    },
    {
      "epoch": 9.323513658275308,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8722,
      "step": 39166
    },
    {
      "epoch": 9.323751711003988,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 39167
    },
    {
      "epoch": 9.323989763732667,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 39168
    },
    {
      "epoch": 9.324227816461347,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9539,
      "step": 39169
    },
    {
      "epoch": 9.324465869190025,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9274,
      "step": 39170
    },
    {
      "epoch": 9.324703921918704,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 39171
    },
    {
      "epoch": 9.324941974647384,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9067,
      "step": 39172
    },
    {
      "epoch": 9.325180027376064,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0132,
      "step": 39173
    },
    {
      "epoch": 9.325418080104743,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0201,
      "step": 39174
    },
    {
      "epoch": 9.325656132833423,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 39175
    },
    {
      "epoch": 9.325894185562102,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9919,
      "step": 39176
    },
    {
      "epoch": 9.326132238290782,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 39177
    },
    {
      "epoch": 9.326370291019462,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8664,
      "step": 39178
    },
    {
      "epoch": 9.32660834374814,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 39179
    },
    {
      "epoch": 9.32684639647682,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 39180
    },
    {
      "epoch": 9.327084449205499,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 39181
    },
    {
      "epoch": 9.327322501934178,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2005,
      "step": 39182
    },
    {
      "epoch": 9.327560554662858,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0741,
      "step": 39183
    },
    {
      "epoch": 9.327798607391538,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3337,
      "step": 39184
    },
    {
      "epoch": 9.328036660120217,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0352,
      "step": 39185
    },
    {
      "epoch": 9.328274712848897,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9212,
      "step": 39186
    },
    {
      "epoch": 9.328512765577575,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 2.0762,
      "step": 39187
    },
    {
      "epoch": 9.328750818306254,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0026,
      "step": 39188
    },
    {
      "epoch": 9.328988871034934,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1011,
      "step": 39189
    },
    {
      "epoch": 9.329226923763613,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 39190
    },
    {
      "epoch": 9.329464976492293,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 39191
    },
    {
      "epoch": 9.329703029220973,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 39192
    },
    {
      "epoch": 9.329941081949652,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0123,
      "step": 39193
    },
    {
      "epoch": 9.330179134678332,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1155,
      "step": 39194
    },
    {
      "epoch": 9.330417187407011,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 39195
    },
    {
      "epoch": 9.33065524013569,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2873,
      "step": 39196
    },
    {
      "epoch": 9.330893292864369,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2455,
      "step": 39197
    },
    {
      "epoch": 9.331131345593048,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.9205,
      "step": 39198
    },
    {
      "epoch": 9.331369398321728,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7904,
      "step": 39199
    },
    {
      "epoch": 9.331607451050408,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2913,
      "step": 39200
    },
    {
      "epoch": 9.331607451050408,
      "eval_runtime": 0.5129,
      "eval_samples_per_second": 1.95,
      "eval_steps_per_second": 1.95,
      "step": 39200
    },
    {
      "epoch": 9.331845503779087,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 39201
    },
    {
      "epoch": 9.332083556507767,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 39202
    },
    {
      "epoch": 9.332321609236446,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9513,
      "step": 39203
    },
    {
      "epoch": 9.332559661965126,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9872,
      "step": 39204
    },
    {
      "epoch": 9.332797714693804,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1628,
      "step": 39205
    },
    {
      "epoch": 9.333035767422484,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8442,
      "step": 39206
    },
    {
      "epoch": 9.333273820151163,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1114,
      "step": 39207
    },
    {
      "epoch": 9.333511872879843,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 39208
    },
    {
      "epoch": 9.333749925608522,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0904,
      "step": 39209
    },
    {
      "epoch": 9.333987978337202,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.301,
      "step": 39210
    },
    {
      "epoch": 9.334226031065882,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7953,
      "step": 39211
    },
    {
      "epoch": 9.334464083794561,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7648,
      "step": 39212
    },
    {
      "epoch": 9.33470213652324,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9371,
      "step": 39213
    },
    {
      "epoch": 9.334940189251919,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9431,
      "step": 39214
    },
    {
      "epoch": 9.335178241980598,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 39215
    },
    {
      "epoch": 9.335416294709278,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1783,
      "step": 39216
    },
    {
      "epoch": 9.335654347437957,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.796,
      "step": 39217
    },
    {
      "epoch": 9.335892400166637,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1687,
      "step": 39218
    },
    {
      "epoch": 9.336130452895317,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 39219
    },
    {
      "epoch": 9.336368505623996,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 39220
    },
    {
      "epoch": 9.336606558352676,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9019,
      "step": 39221
    },
    {
      "epoch": 9.336844611081354,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 39222
    },
    {
      "epoch": 9.337082663810033,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 39223
    },
    {
      "epoch": 9.337320716538713,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.957,
      "step": 39224
    },
    {
      "epoch": 9.337558769267392,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7624,
      "step": 39225
    },
    {
      "epoch": 9.337796821996072,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2032,
      "step": 39226
    },
    {
      "epoch": 9.338034874724752,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 39227
    },
    {
      "epoch": 9.338272927453431,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 39228
    },
    {
      "epoch": 9.338510980182111,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9727,
      "step": 39229
    },
    {
      "epoch": 9.33874903291079,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0606,
      "step": 39230
    },
    {
      "epoch": 9.338987085639468,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 39231
    },
    {
      "epoch": 9.339225138368148,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.7865,
      "step": 39232
    },
    {
      "epoch": 9.339463191096828,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0045,
      "step": 39233
    },
    {
      "epoch": 9.339701243825507,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3351,
      "step": 39234
    },
    {
      "epoch": 9.339939296554187,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8102,
      "step": 39235
    },
    {
      "epoch": 9.340177349282866,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9126,
      "step": 39236
    },
    {
      "epoch": 9.340415402011546,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 39237
    },
    {
      "epoch": 9.340653454740226,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9684,
      "step": 39238
    },
    {
      "epoch": 9.340891507468905,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2732,
      "step": 39239
    },
    {
      "epoch": 9.341129560197583,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9959,
      "step": 39240
    },
    {
      "epoch": 9.341367612926263,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9666,
      "step": 39241
    },
    {
      "epoch": 9.341605665654942,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0091,
      "step": 39242
    },
    {
      "epoch": 9.341843718383622,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0574,
      "step": 39243
    },
    {
      "epoch": 9.342081771112301,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9958,
      "step": 39244
    },
    {
      "epoch": 9.342319823840981,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.0846,
      "step": 39245
    },
    {
      "epoch": 9.34255787656966,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2676,
      "step": 39246
    },
    {
      "epoch": 9.34279592929834,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9049,
      "step": 39247
    },
    {
      "epoch": 9.34303398202702,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0373,
      "step": 39248
    },
    {
      "epoch": 9.343272034755698,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 39249
    },
    {
      "epoch": 9.343510087484377,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 39250
    },
    {
      "epoch": 9.343748140213057,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0388,
      "step": 39251
    },
    {
      "epoch": 9.343986192941736,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3948,
      "step": 39252
    },
    {
      "epoch": 9.344224245670416,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 39253
    },
    {
      "epoch": 9.344462298399096,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 39254
    },
    {
      "epoch": 9.344700351127775,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 39255
    },
    {
      "epoch": 9.344938403856455,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 39256
    },
    {
      "epoch": 9.345176456585133,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 39257
    },
    {
      "epoch": 9.345414509313812,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 39258
    },
    {
      "epoch": 9.345652562042492,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.2089,
      "step": 39259
    },
    {
      "epoch": 9.345890614771172,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2059,
      "step": 39260
    },
    {
      "epoch": 9.346128667499851,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9507,
      "step": 39261
    },
    {
      "epoch": 9.34636672022853,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0277,
      "step": 39262
    },
    {
      "epoch": 9.34660477295721,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 39263
    },
    {
      "epoch": 9.34684282568589,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 39264
    },
    {
      "epoch": 9.34708087841457,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 39265
    },
    {
      "epoch": 9.347318931143247,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.2959,
      "step": 39266
    },
    {
      "epoch": 9.347556983871927,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 39267
    },
    {
      "epoch": 9.347795036600607,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9484,
      "step": 39268
    },
    {
      "epoch": 9.348033089329286,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9196,
      "step": 39269
    },
    {
      "epoch": 9.348271142057966,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.5274,
      "step": 39270
    },
    {
      "epoch": 9.348509194786645,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 39271
    },
    {
      "epoch": 9.348747247515325,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0471,
      "step": 39272
    },
    {
      "epoch": 9.348985300244005,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 39273
    },
    {
      "epoch": 9.349223352972684,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.3128,
      "step": 39274
    },
    {
      "epoch": 9.349461405701362,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 39275
    },
    {
      "epoch": 9.349699458430042,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9511,
      "step": 39276
    },
    {
      "epoch": 9.349937511158721,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 39277
    },
    {
      "epoch": 9.350175563887401,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 39278
    },
    {
      "epoch": 9.35041361661608,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1122,
      "step": 39279
    },
    {
      "epoch": 9.35065166934476,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1236,
      "step": 39280
    },
    {
      "epoch": 9.35088972207344,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 39281
    },
    {
      "epoch": 9.35112777480212,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1391,
      "step": 39282
    },
    {
      "epoch": 9.351365827530799,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 39283
    },
    {
      "epoch": 9.351603880259477,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 39284
    },
    {
      "epoch": 9.351841932988156,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 39285
    },
    {
      "epoch": 9.352079985716836,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9369,
      "step": 39286
    },
    {
      "epoch": 9.352318038445516,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 39287
    },
    {
      "epoch": 9.352556091174195,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 39288
    },
    {
      "epoch": 9.352794143902875,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8202,
      "step": 39289
    },
    {
      "epoch": 9.353032196631554,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1589,
      "step": 39290
    },
    {
      "epoch": 9.353270249360234,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.906,
      "step": 39291
    },
    {
      "epoch": 9.353508302088912,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0819,
      "step": 39292
    },
    {
      "epoch": 9.353746354817591,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 39293
    },
    {
      "epoch": 9.353984407546271,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 39294
    },
    {
      "epoch": 9.35422246027495,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 39295
    },
    {
      "epoch": 9.35446051300363,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7639,
      "step": 39296
    },
    {
      "epoch": 9.35469856573231,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9921,
      "step": 39297
    },
    {
      "epoch": 9.35493661846099,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4619,
      "step": 39298
    },
    {
      "epoch": 9.355174671189669,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.005,
      "step": 39299
    },
    {
      "epoch": 9.355412723918349,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1314,
      "step": 39300
    },
    {
      "epoch": 9.355412723918349,
      "eval_runtime": 0.5256,
      "eval_samples_per_second": 1.903,
      "eval_steps_per_second": 1.903,
      "step": 39300
    },
    {
      "epoch": 9.355650776647026,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2345,
      "step": 39301
    },
    {
      "epoch": 9.355888829375706,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 39302
    },
    {
      "epoch": 9.356126882104386,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9931,
      "step": 39303
    },
    {
      "epoch": 9.356364934833065,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.098,
      "step": 39304
    },
    {
      "epoch": 9.356602987561745,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 39305
    },
    {
      "epoch": 9.356841040290425,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9691,
      "step": 39306
    },
    {
      "epoch": 9.357079093019104,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2049,
      "step": 39307
    },
    {
      "epoch": 9.357317145747784,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.5994,
      "step": 39308
    },
    {
      "epoch": 9.357555198476463,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.6694,
      "step": 39309
    },
    {
      "epoch": 9.357793251205141,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0989,
      "step": 39310
    },
    {
      "epoch": 9.35803130393382,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2027,
      "step": 39311
    },
    {
      "epoch": 9.3582693566625,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1774,
      "step": 39312
    },
    {
      "epoch": 9.35850740939118,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0049,
      "step": 39313
    },
    {
      "epoch": 9.35874546211986,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2475,
      "step": 39314
    },
    {
      "epoch": 9.35898351484854,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.231,
      "step": 39315
    },
    {
      "epoch": 9.359221567577219,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9241,
      "step": 39316
    },
    {
      "epoch": 9.359459620305898,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3085,
      "step": 39317
    },
    {
      "epoch": 9.359697673034578,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 39318
    },
    {
      "epoch": 9.359935725763256,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0143,
      "step": 39319
    },
    {
      "epoch": 9.360173778491935,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 39320
    },
    {
      "epoch": 9.360411831220615,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 39321
    },
    {
      "epoch": 9.360649883949295,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.243,
      "step": 39322
    },
    {
      "epoch": 9.360887936677974,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1143,
      "step": 39323
    },
    {
      "epoch": 9.361125989406654,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 39324
    },
    {
      "epoch": 9.361364042135333,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0144,
      "step": 39325
    },
    {
      "epoch": 9.361602094864013,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7289,
      "step": 39326
    },
    {
      "epoch": 9.361840147592691,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.6433,
      "step": 39327
    },
    {
      "epoch": 9.36207820032137,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.997,
      "step": 39328
    },
    {
      "epoch": 9.36231625305005,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 39329
    },
    {
      "epoch": 9.36255430577873,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.7308,
      "step": 39330
    },
    {
      "epoch": 9.36279235850741,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 39331
    },
    {
      "epoch": 9.363030411236089,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0038,
      "step": 39332
    },
    {
      "epoch": 9.363268463964769,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 39333
    },
    {
      "epoch": 9.363506516693448,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 39334
    },
    {
      "epoch": 9.363744569422128,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0737,
      "step": 39335
    },
    {
      "epoch": 9.363982622150806,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0318,
      "step": 39336
    },
    {
      "epoch": 9.364220674879485,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 39337
    },
    {
      "epoch": 9.364458727608165,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 39338
    },
    {
      "epoch": 9.364696780336844,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0597,
      "step": 39339
    },
    {
      "epoch": 9.364934833065524,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.085,
      "step": 39340
    },
    {
      "epoch": 9.365172885794204,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.3523,
      "step": 39341
    },
    {
      "epoch": 9.365410938522883,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.29,
      "step": 39342
    },
    {
      "epoch": 9.365648991251563,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.4727,
      "step": 39343
    },
    {
      "epoch": 9.365887043980242,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8976,
      "step": 39344
    },
    {
      "epoch": 9.36612509670892,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8834,
      "step": 39345
    },
    {
      "epoch": 9.3663631494376,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9562,
      "step": 39346
    },
    {
      "epoch": 9.36660120216628,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3484,
      "step": 39347
    },
    {
      "epoch": 9.366839254894959,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1638,
      "step": 39348
    },
    {
      "epoch": 9.367077307623639,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 39349
    },
    {
      "epoch": 9.367315360352318,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1692,
      "step": 39350
    },
    {
      "epoch": 9.367553413080998,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.7681,
      "step": 39351
    },
    {
      "epoch": 9.367791465809677,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8707,
      "step": 39352
    },
    {
      "epoch": 9.368029518538357,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9298,
      "step": 39353
    },
    {
      "epoch": 9.368267571267035,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3038,
      "step": 39354
    },
    {
      "epoch": 9.368505623995715,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 39355
    },
    {
      "epoch": 9.368743676724394,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 39356
    },
    {
      "epoch": 9.368981729453074,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 39357
    },
    {
      "epoch": 9.369219782181753,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 39358
    },
    {
      "epoch": 9.369457834910433,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9504,
      "step": 39359
    },
    {
      "epoch": 9.369695887639113,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 39360
    },
    {
      "epoch": 9.369933940367792,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.7991,
      "step": 39361
    },
    {
      "epoch": 9.37017199309647,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8,
      "step": 39362
    },
    {
      "epoch": 9.37041004582515,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 39363
    },
    {
      "epoch": 9.37064809855383,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8754,
      "step": 39364
    },
    {
      "epoch": 9.370886151282509,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9861,
      "step": 39365
    },
    {
      "epoch": 9.371124204011188,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.847,
      "step": 39366
    },
    {
      "epoch": 9.371362256739868,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9155,
      "step": 39367
    },
    {
      "epoch": 9.371600309468548,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7866,
      "step": 39368
    },
    {
      "epoch": 9.371838362197227,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0533,
      "step": 39369
    },
    {
      "epoch": 9.372076414925907,
      "grad_norm": 7.375,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 39370
    },
    {
      "epoch": 9.372314467654585,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 39371
    },
    {
      "epoch": 9.372552520383264,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.091,
      "step": 39372
    },
    {
      "epoch": 9.372790573111944,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 39373
    },
    {
      "epoch": 9.373028625840623,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0034,
      "step": 39374
    },
    {
      "epoch": 9.373266678569303,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 39375
    },
    {
      "epoch": 9.373504731297983,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 39376
    },
    {
      "epoch": 9.373742784026662,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 39377
    },
    {
      "epoch": 9.373980836755342,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.7312,
      "step": 39378
    },
    {
      "epoch": 9.374218889484021,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9222,
      "step": 39379
    },
    {
      "epoch": 9.3744569422127,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 39380
    },
    {
      "epoch": 9.374694994941379,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9694,
      "step": 39381
    },
    {
      "epoch": 9.374933047670059,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 39382
    },
    {
      "epoch": 9.375171100398738,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 39383
    },
    {
      "epoch": 9.375409153127418,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.1928,
      "step": 39384
    },
    {
      "epoch": 9.375647205856097,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7976,
      "step": 39385
    },
    {
      "epoch": 9.375885258584777,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.3544,
      "step": 39386
    },
    {
      "epoch": 9.376123311313457,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0193,
      "step": 39387
    },
    {
      "epoch": 9.376361364042136,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8285,
      "step": 39388
    },
    {
      "epoch": 9.376599416770814,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9713,
      "step": 39389
    },
    {
      "epoch": 9.376837469499494,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 39390
    },
    {
      "epoch": 9.377075522228173,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0773,
      "step": 39391
    },
    {
      "epoch": 9.377313574956853,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 39392
    },
    {
      "epoch": 9.377551627685532,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2981,
      "step": 39393
    },
    {
      "epoch": 9.377789680414212,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.901,
      "step": 39394
    },
    {
      "epoch": 9.378027733142892,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9545,
      "step": 39395
    },
    {
      "epoch": 9.378265785871571,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9568,
      "step": 39396
    },
    {
      "epoch": 9.378503838600249,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9455,
      "step": 39397
    },
    {
      "epoch": 9.378741891328929,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0692,
      "step": 39398
    },
    {
      "epoch": 9.378979944057608,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0794,
      "step": 39399
    },
    {
      "epoch": 9.379217996786288,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 39400
    },
    {
      "epoch": 9.379217996786288,
      "eval_runtime": 0.5204,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 39400
    },
    {
      "epoch": 9.379456049514967,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2677,
      "step": 39401
    },
    {
      "epoch": 9.379694102243647,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0154,
      "step": 39402
    },
    {
      "epoch": 9.379932154972327,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 39403
    },
    {
      "epoch": 9.380170207701006,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.6976,
      "step": 39404
    },
    {
      "epoch": 9.380408260429686,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 39405
    },
    {
      "epoch": 9.380646313158364,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 39406
    },
    {
      "epoch": 9.380884365887043,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1536,
      "step": 39407
    },
    {
      "epoch": 9.381122418615723,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1375,
      "step": 39408
    },
    {
      "epoch": 9.381360471344403,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0269,
      "step": 39409
    },
    {
      "epoch": 9.381598524073082,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2562,
      "step": 39410
    },
    {
      "epoch": 9.381836576801762,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9577,
      "step": 39411
    },
    {
      "epoch": 9.382074629530441,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 39412
    },
    {
      "epoch": 9.382312682259121,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1552,
      "step": 39413
    },
    {
      "epoch": 9.3825507349878,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 39414
    },
    {
      "epoch": 9.382788787716478,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1796,
      "step": 39415
    },
    {
      "epoch": 9.383026840445158,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 39416
    },
    {
      "epoch": 9.383264893173838,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 39417
    },
    {
      "epoch": 9.383502945902517,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.7657,
      "step": 39418
    },
    {
      "epoch": 9.383740998631197,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 39419
    },
    {
      "epoch": 9.383979051359876,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2949,
      "step": 39420
    },
    {
      "epoch": 9.384217104088556,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0594,
      "step": 39421
    },
    {
      "epoch": 9.384455156817236,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7856,
      "step": 39422
    },
    {
      "epoch": 9.384693209545915,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1535,
      "step": 39423
    },
    {
      "epoch": 9.384931262274593,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2699,
      "step": 39424
    },
    {
      "epoch": 9.385169315003273,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8906,
      "step": 39425
    },
    {
      "epoch": 9.385407367731952,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 39426
    },
    {
      "epoch": 9.385645420460632,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0894,
      "step": 39427
    },
    {
      "epoch": 9.385883473189311,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.131,
      "step": 39428
    },
    {
      "epoch": 9.386121525917991,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 39429
    },
    {
      "epoch": 9.38635957864667,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 39430
    },
    {
      "epoch": 9.38659763137535,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7982,
      "step": 39431
    },
    {
      "epoch": 9.386835684104028,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 39432
    },
    {
      "epoch": 9.387073736832708,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 39433
    },
    {
      "epoch": 9.387311789561387,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1444,
      "step": 39434
    },
    {
      "epoch": 9.387549842290067,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9264,
      "step": 39435
    },
    {
      "epoch": 9.387787895018747,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0616,
      "step": 39436
    },
    {
      "epoch": 9.388025947747426,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8635,
      "step": 39437
    },
    {
      "epoch": 9.388264000476106,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0137,
      "step": 39438
    },
    {
      "epoch": 9.388502053204785,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 39439
    },
    {
      "epoch": 9.388740105933465,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 39440
    },
    {
      "epoch": 9.388978158662145,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.627,
      "step": 39441
    },
    {
      "epoch": 9.389216211390822,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2436,
      "step": 39442
    },
    {
      "epoch": 9.389454264119502,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9767,
      "step": 39443
    },
    {
      "epoch": 9.389692316848182,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 39444
    },
    {
      "epoch": 9.389930369576861,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 39445
    },
    {
      "epoch": 9.39016842230554,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 39446
    },
    {
      "epoch": 9.39040647503422,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8121,
      "step": 39447
    },
    {
      "epoch": 9.3906445277629,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2449,
      "step": 39448
    },
    {
      "epoch": 9.39088258049158,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 39449
    },
    {
      "epoch": 9.391120633220257,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 39450
    },
    {
      "epoch": 9.391358685948937,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1607,
      "step": 39451
    },
    {
      "epoch": 9.391596738677617,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7605,
      "step": 39452
    },
    {
      "epoch": 9.391834791406296,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 39453
    },
    {
      "epoch": 9.392072844134976,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.833,
      "step": 39454
    },
    {
      "epoch": 9.392310896863656,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2198,
      "step": 39455
    },
    {
      "epoch": 9.392548949592335,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 39456
    },
    {
      "epoch": 9.392787002321015,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0839,
      "step": 39457
    },
    {
      "epoch": 9.393025055049694,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 39458
    },
    {
      "epoch": 9.393263107778372,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1028,
      "step": 39459
    },
    {
      "epoch": 9.393501160507052,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0039,
      "step": 39460
    },
    {
      "epoch": 9.393739213235731,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.0376,
      "step": 39461
    },
    {
      "epoch": 9.393977265964411,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 39462
    },
    {
      "epoch": 9.39421531869309,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9965,
      "step": 39463
    },
    {
      "epoch": 9.39445337142177,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 39464
    },
    {
      "epoch": 9.39469142415045,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 39465
    },
    {
      "epoch": 9.39492947687913,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9934,
      "step": 39466
    },
    {
      "epoch": 9.395167529607809,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.959,
      "step": 39467
    },
    {
      "epoch": 9.395405582336487,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.983,
      "step": 39468
    },
    {
      "epoch": 9.395643635065166,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9368,
      "step": 39469
    },
    {
      "epoch": 9.395881687793846,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0456,
      "step": 39470
    },
    {
      "epoch": 9.396119740522526,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0037,
      "step": 39471
    },
    {
      "epoch": 9.396357793251205,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1116,
      "step": 39472
    },
    {
      "epoch": 9.396595845979885,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 39473
    },
    {
      "epoch": 9.396833898708564,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8555,
      "step": 39474
    },
    {
      "epoch": 9.397071951437244,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1267,
      "step": 39475
    },
    {
      "epoch": 9.397310004165924,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0926,
      "step": 39476
    },
    {
      "epoch": 9.397548056894601,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1448,
      "step": 39477
    },
    {
      "epoch": 9.397786109623281,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9707,
      "step": 39478
    },
    {
      "epoch": 9.39802416235196,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 39479
    },
    {
      "epoch": 9.39826221508064,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2199,
      "step": 39480
    },
    {
      "epoch": 9.39850026780932,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.409,
      "step": 39481
    },
    {
      "epoch": 9.398738320538,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.1801,
      "step": 39482
    },
    {
      "epoch": 9.39897637326668,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 39483
    },
    {
      "epoch": 9.399214425995359,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 39484
    },
    {
      "epoch": 9.399452478724037,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9064,
      "step": 39485
    },
    {
      "epoch": 9.399690531452716,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1874,
      "step": 39486
    },
    {
      "epoch": 9.399928584181396,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9057,
      "step": 39487
    },
    {
      "epoch": 9.400166636910075,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9836,
      "step": 39488
    },
    {
      "epoch": 9.400404689638755,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7776,
      "step": 39489
    },
    {
      "epoch": 9.400642742367435,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8984,
      "step": 39490
    },
    {
      "epoch": 9.400880795096114,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0662,
      "step": 39491
    },
    {
      "epoch": 9.401118847824794,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8689,
      "step": 39492
    },
    {
      "epoch": 9.401356900553473,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 39493
    },
    {
      "epoch": 9.401594953282151,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.5379,
      "step": 39494
    },
    {
      "epoch": 9.40183300601083,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 39495
    },
    {
      "epoch": 9.40207105873951,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8544,
      "step": 39496
    },
    {
      "epoch": 9.40230911146819,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2195,
      "step": 39497
    },
    {
      "epoch": 9.40254716419687,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 39498
    },
    {
      "epoch": 9.40278521692555,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0265,
      "step": 39499
    },
    {
      "epoch": 9.403023269654229,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8276,
      "step": 39500
    },
    {
      "epoch": 9.403023269654229,
      "eval_runtime": 0.5199,
      "eval_samples_per_second": 1.923,
      "eval_steps_per_second": 1.923,
      "step": 39500
    },
    {
      "epoch": 9.403261322382908,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.918,
      "step": 39501
    },
    {
      "epoch": 9.403499375111588,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2945,
      "step": 39502
    },
    {
      "epoch": 9.403737427840266,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9027,
      "step": 39503
    },
    {
      "epoch": 9.403975480568945,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0771,
      "step": 39504
    },
    {
      "epoch": 9.404213533297625,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 39505
    },
    {
      "epoch": 9.404451586026305,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9649,
      "step": 39506
    },
    {
      "epoch": 9.404689638754984,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1722,
      "step": 39507
    },
    {
      "epoch": 9.404927691483664,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3063,
      "step": 39508
    },
    {
      "epoch": 9.405165744212344,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8407,
      "step": 39509
    },
    {
      "epoch": 9.405403796941023,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0635,
      "step": 39510
    },
    {
      "epoch": 9.405641849669703,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2768,
      "step": 39511
    },
    {
      "epoch": 9.40587990239838,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1616,
      "step": 39512
    },
    {
      "epoch": 9.40611795512706,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9957,
      "step": 39513
    },
    {
      "epoch": 9.40635600785574,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 2.0105,
      "step": 39514
    },
    {
      "epoch": 9.40659406058442,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9656,
      "step": 39515
    },
    {
      "epoch": 9.406832113313099,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9554,
      "step": 39516
    },
    {
      "epoch": 9.407070166041779,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.925,
      "step": 39517
    },
    {
      "epoch": 9.407308218770458,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3741,
      "step": 39518
    },
    {
      "epoch": 9.407546271499138,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0172,
      "step": 39519
    },
    {
      "epoch": 9.407784324227816,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1555,
      "step": 39520
    },
    {
      "epoch": 9.408022376956495,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0983,
      "step": 39521
    },
    {
      "epoch": 9.408260429685175,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2321,
      "step": 39522
    },
    {
      "epoch": 9.408498482413854,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0972,
      "step": 39523
    },
    {
      "epoch": 9.408736535142534,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 39524
    },
    {
      "epoch": 9.408974587871214,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.3528,
      "step": 39525
    },
    {
      "epoch": 9.409212640599893,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.786,
      "step": 39526
    },
    {
      "epoch": 9.409450693328573,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0916,
      "step": 39527
    },
    {
      "epoch": 9.409688746057252,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1726,
      "step": 39528
    },
    {
      "epoch": 9.40992679878593,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.88,
      "step": 39529
    },
    {
      "epoch": 9.41016485151461,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2215,
      "step": 39530
    },
    {
      "epoch": 9.41040290424329,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 39531
    },
    {
      "epoch": 9.41064095697197,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0693,
      "step": 39532
    },
    {
      "epoch": 9.410879009700649,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9766,
      "step": 39533
    },
    {
      "epoch": 9.411117062429328,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.2057,
      "step": 39534
    },
    {
      "epoch": 9.411355115158008,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3071,
      "step": 39535
    },
    {
      "epoch": 9.411593167886688,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 39536
    },
    {
      "epoch": 9.411831220615367,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 39537
    },
    {
      "epoch": 9.412069273344045,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 39538
    },
    {
      "epoch": 9.412307326072725,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1583,
      "step": 39539
    },
    {
      "epoch": 9.412545378801404,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1338,
      "step": 39540
    },
    {
      "epoch": 9.412783431530084,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 39541
    },
    {
      "epoch": 9.413021484258763,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.7336,
      "step": 39542
    },
    {
      "epoch": 9.413259536987443,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0885,
      "step": 39543
    },
    {
      "epoch": 9.413497589716123,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 39544
    },
    {
      "epoch": 9.413735642444802,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 39545
    },
    {
      "epoch": 9.413973695173482,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9276,
      "step": 39546
    },
    {
      "epoch": 9.41421174790216,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.1675,
      "step": 39547
    },
    {
      "epoch": 9.41444980063084,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3629,
      "step": 39548
    },
    {
      "epoch": 9.414687853359519,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.262,
      "step": 39549
    },
    {
      "epoch": 9.414925906088198,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9239,
      "step": 39550
    },
    {
      "epoch": 9.415163958816878,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0149,
      "step": 39551
    },
    {
      "epoch": 9.415402011545558,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 39552
    },
    {
      "epoch": 9.415640064274237,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 39553
    },
    {
      "epoch": 9.415878117002917,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 39554
    },
    {
      "epoch": 9.416116169731595,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 39555
    },
    {
      "epoch": 9.416354222460274,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9257,
      "step": 39556
    },
    {
      "epoch": 9.416592275188954,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9248,
      "step": 39557
    },
    {
      "epoch": 9.416830327917634,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0126,
      "step": 39558
    },
    {
      "epoch": 9.417068380646313,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8565,
      "step": 39559
    },
    {
      "epoch": 9.417306433374993,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.7173,
      "step": 39560
    },
    {
      "epoch": 9.417544486103672,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1003,
      "step": 39561
    },
    {
      "epoch": 9.417782538832352,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0255,
      "step": 39562
    },
    {
      "epoch": 9.418020591561032,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1171,
      "step": 39563
    },
    {
      "epoch": 9.41825864428971,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9652,
      "step": 39564
    },
    {
      "epoch": 9.418496697018389,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9816,
      "step": 39565
    },
    {
      "epoch": 9.418734749747069,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0955,
      "step": 39566
    },
    {
      "epoch": 9.418972802475748,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0307,
      "step": 39567
    },
    {
      "epoch": 9.419210855204428,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 39568
    },
    {
      "epoch": 9.419448907933107,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.4059,
      "step": 39569
    },
    {
      "epoch": 9.419686960661787,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 39570
    },
    {
      "epoch": 9.419925013390467,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9157,
      "step": 39571
    },
    {
      "epoch": 9.420163066119146,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0476,
      "step": 39572
    },
    {
      "epoch": 9.420401118847824,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 39573
    },
    {
      "epoch": 9.420639171576504,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.947,
      "step": 39574
    },
    {
      "epoch": 9.420877224305183,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.3094,
      "step": 39575
    },
    {
      "epoch": 9.421115277033863,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0805,
      "step": 39576
    },
    {
      "epoch": 9.421353329762542,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.911,
      "step": 39577
    },
    {
      "epoch": 9.421591382491222,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 39578
    },
    {
      "epoch": 9.421829435219902,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2902,
      "step": 39579
    },
    {
      "epoch": 9.422067487948581,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7557,
      "step": 39580
    },
    {
      "epoch": 9.422305540677261,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.3934,
      "step": 39581
    },
    {
      "epoch": 9.422543593405939,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1921,
      "step": 39582
    },
    {
      "epoch": 9.422781646134618,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.001,
      "step": 39583
    },
    {
      "epoch": 9.423019698863298,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.8946,
      "step": 39584
    },
    {
      "epoch": 9.423257751591978,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 39585
    },
    {
      "epoch": 9.423495804320657,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0454,
      "step": 39586
    },
    {
      "epoch": 9.423733857049337,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8148,
      "step": 39587
    },
    {
      "epoch": 9.423971909778016,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1871,
      "step": 39588
    },
    {
      "epoch": 9.424209962506696,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1741,
      "step": 39589
    },
    {
      "epoch": 9.424448015235374,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0562,
      "step": 39590
    },
    {
      "epoch": 9.424686067964053,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0851,
      "step": 39591
    },
    {
      "epoch": 9.424924120692733,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1315,
      "step": 39592
    },
    {
      "epoch": 9.425162173421413,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.822,
      "step": 39593
    },
    {
      "epoch": 9.425400226150092,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0951,
      "step": 39594
    },
    {
      "epoch": 9.425638278878772,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 39595
    },
    {
      "epoch": 9.425876331607451,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8875,
      "step": 39596
    },
    {
      "epoch": 9.426114384336131,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3695,
      "step": 39597
    },
    {
      "epoch": 9.42635243706481,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9824,
      "step": 39598
    },
    {
      "epoch": 9.426590489793488,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.0586,
      "step": 39599
    },
    {
      "epoch": 9.426828542522168,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.907,
      "step": 39600
    },
    {
      "epoch": 9.426828542522168,
      "eval_runtime": 0.5202,
      "eval_samples_per_second": 1.922,
      "eval_steps_per_second": 1.922,
      "step": 39600
    },
    {
      "epoch": 9.427066595250848,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 39601
    },
    {
      "epoch": 9.427304647979527,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1044,
      "step": 39602
    },
    {
      "epoch": 9.427542700708207,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8266,
      "step": 39603
    },
    {
      "epoch": 9.427780753436886,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8806,
      "step": 39604
    },
    {
      "epoch": 9.428018806165566,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 39605
    },
    {
      "epoch": 9.428256858894246,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.2806,
      "step": 39606
    },
    {
      "epoch": 9.428494911622925,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8076,
      "step": 39607
    },
    {
      "epoch": 9.428732964351603,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1117,
      "step": 39608
    },
    {
      "epoch": 9.428971017080283,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 39609
    },
    {
      "epoch": 9.429209069808962,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0565,
      "step": 39610
    },
    {
      "epoch": 9.429447122537642,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1333,
      "step": 39611
    },
    {
      "epoch": 9.429685175266322,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 39612
    },
    {
      "epoch": 9.429923227995001,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.81,
      "step": 39613
    },
    {
      "epoch": 9.43016128072368,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 39614
    },
    {
      "epoch": 9.43039933345236,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8012,
      "step": 39615
    },
    {
      "epoch": 9.43063738618104,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2093,
      "step": 39616
    },
    {
      "epoch": 9.430875438909718,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0918,
      "step": 39617
    },
    {
      "epoch": 9.431113491638397,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.022,
      "step": 39618
    },
    {
      "epoch": 9.431351544367077,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1088,
      "step": 39619
    },
    {
      "epoch": 9.431589597095757,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 39620
    },
    {
      "epoch": 9.431827649824436,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0314,
      "step": 39621
    },
    {
      "epoch": 9.432065702553116,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 39622
    },
    {
      "epoch": 9.432303755281795,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 39623
    },
    {
      "epoch": 9.432541808010475,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9847,
      "step": 39624
    },
    {
      "epoch": 9.432779860739153,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 39625
    },
    {
      "epoch": 9.433017913467832,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6322,
      "step": 39626
    },
    {
      "epoch": 9.433255966196512,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8714,
      "step": 39627
    },
    {
      "epoch": 9.433494018925192,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1643,
      "step": 39628
    },
    {
      "epoch": 9.433732071653871,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0249,
      "step": 39629
    },
    {
      "epoch": 9.433970124382551,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.2548,
      "step": 39630
    },
    {
      "epoch": 9.43420817711123,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0798,
      "step": 39631
    },
    {
      "epoch": 9.43444622983991,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.7788,
      "step": 39632
    },
    {
      "epoch": 9.43468428256859,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9402,
      "step": 39633
    },
    {
      "epoch": 9.434922335297268,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9613,
      "step": 39634
    },
    {
      "epoch": 9.435160388025947,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0315,
      "step": 39635
    },
    {
      "epoch": 9.435398440754627,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1436,
      "step": 39636
    },
    {
      "epoch": 9.435636493483306,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 39637
    },
    {
      "epoch": 9.435874546211986,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8731,
      "step": 39638
    },
    {
      "epoch": 9.436112598940666,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2238,
      "step": 39639
    },
    {
      "epoch": 9.436350651669345,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9869,
      "step": 39640
    },
    {
      "epoch": 9.436588704398025,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3013,
      "step": 39641
    },
    {
      "epoch": 9.436826757126704,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9372,
      "step": 39642
    },
    {
      "epoch": 9.437064809855382,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8259,
      "step": 39643
    },
    {
      "epoch": 9.437302862584062,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 39644
    },
    {
      "epoch": 9.437540915312741,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0223,
      "step": 39645
    },
    {
      "epoch": 9.437778968041421,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8349,
      "step": 39646
    },
    {
      "epoch": 9.4380170207701,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9497,
      "step": 39647
    },
    {
      "epoch": 9.43825507349878,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8999,
      "step": 39648
    },
    {
      "epoch": 9.43849312622746,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 39649
    },
    {
      "epoch": 9.43873117895614,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2482,
      "step": 39650
    },
    {
      "epoch": 9.438969231684819,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7957,
      "step": 39651
    },
    {
      "epoch": 9.439207284413497,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8873,
      "step": 39652
    },
    {
      "epoch": 9.439445337142176,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0598,
      "step": 39653
    },
    {
      "epoch": 9.439683389870856,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1193,
      "step": 39654
    },
    {
      "epoch": 9.439921442599536,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.0345,
      "step": 39655
    },
    {
      "epoch": 9.440159495328215,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2991,
      "step": 39656
    },
    {
      "epoch": 9.440397548056895,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 39657
    },
    {
      "epoch": 9.440635600785575,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2122,
      "step": 39658
    },
    {
      "epoch": 9.440873653514254,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0296,
      "step": 39659
    },
    {
      "epoch": 9.441111706242932,
      "grad_norm": 5.84375,
      "learning_rate": 3e-05,
      "loss": 2.1659,
      "step": 39660
    },
    {
      "epoch": 9.441349758971612,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0697,
      "step": 39661
    },
    {
      "epoch": 9.441587811700291,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.4465,
      "step": 39662
    },
    {
      "epoch": 9.44182586442897,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.967,
      "step": 39663
    },
    {
      "epoch": 9.44206391715765,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2909,
      "step": 39664
    },
    {
      "epoch": 9.44230196988633,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.142,
      "step": 39665
    },
    {
      "epoch": 9.44254002261501,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0855,
      "step": 39666
    },
    {
      "epoch": 9.44277807534369,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9315,
      "step": 39667
    },
    {
      "epoch": 9.443016128072369,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.3075,
      "step": 39668
    },
    {
      "epoch": 9.443254180801047,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9396,
      "step": 39669
    },
    {
      "epoch": 9.443492233529726,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.539,
      "step": 39670
    },
    {
      "epoch": 9.443730286258406,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0027,
      "step": 39671
    },
    {
      "epoch": 9.443968338987085,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 39672
    },
    {
      "epoch": 9.444206391715765,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.1902,
      "step": 39673
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9004,
      "step": 39674
    },
    {
      "epoch": 9.444682497173124,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.7614,
      "step": 39675
    },
    {
      "epoch": 9.444920549901804,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9219,
      "step": 39676
    },
    {
      "epoch": 9.445158602630483,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.991,
      "step": 39677
    },
    {
      "epoch": 9.445396655359161,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 39678
    },
    {
      "epoch": 9.445634708087841,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0443,
      "step": 39679
    },
    {
      "epoch": 9.44587276081652,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1341,
      "step": 39680
    },
    {
      "epoch": 9.4461108135452,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 39681
    },
    {
      "epoch": 9.44634886627388,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 39682
    },
    {
      "epoch": 9.44658691900256,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9754,
      "step": 39683
    },
    {
      "epoch": 9.446824971731239,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1514,
      "step": 39684
    },
    {
      "epoch": 9.447063024459919,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 39685
    },
    {
      "epoch": 9.447301077188598,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.877,
      "step": 39686
    },
    {
      "epoch": 9.447539129917276,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3066,
      "step": 39687
    },
    {
      "epoch": 9.447777182645956,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 39688
    },
    {
      "epoch": 9.448015235374635,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2443,
      "step": 39689
    },
    {
      "epoch": 9.448253288103315,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 39690
    },
    {
      "epoch": 9.448491340831994,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.4523,
      "step": 39691
    },
    {
      "epoch": 9.448729393560674,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0125,
      "step": 39692
    },
    {
      "epoch": 9.448967446289354,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 39693
    },
    {
      "epoch": 9.449205499018033,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0425,
      "step": 39694
    },
    {
      "epoch": 9.449443551746711,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1413,
      "step": 39695
    },
    {
      "epoch": 9.44968160447539,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1414,
      "step": 39696
    },
    {
      "epoch": 9.44991965720407,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0124,
      "step": 39697
    },
    {
      "epoch": 9.45015770993275,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 39698
    },
    {
      "epoch": 9.45039576266143,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1503,
      "step": 39699
    },
    {
      "epoch": 9.450633815390109,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2782,
      "step": 39700
    },
    {
      "epoch": 9.450633815390109,
      "eval_runtime": 0.53,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 39700
    },
    {
      "epoch": 9.450871868118789,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8915,
      "step": 39701
    },
    {
      "epoch": 9.451109920847468,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9418,
      "step": 39702
    },
    {
      "epoch": 9.451347973576148,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0441,
      "step": 39703
    },
    {
      "epoch": 9.451586026304826,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.9499,
      "step": 39704
    },
    {
      "epoch": 9.451824079033505,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8543,
      "step": 39705
    },
    {
      "epoch": 9.452062131762185,
      "grad_norm": 6.34375,
      "learning_rate": 3e-05,
      "loss": 1.7109,
      "step": 39706
    },
    {
      "epoch": 9.452300184490865,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 39707
    },
    {
      "epoch": 9.452538237219544,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8336,
      "step": 39708
    },
    {
      "epoch": 9.452776289948224,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 39709
    },
    {
      "epoch": 9.453014342676903,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0795,
      "step": 39710
    },
    {
      "epoch": 9.453252395405583,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 39711
    },
    {
      "epoch": 9.453490448134263,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 39712
    },
    {
      "epoch": 9.45372850086294,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0552,
      "step": 39713
    },
    {
      "epoch": 9.45396655359162,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8445,
      "step": 39714
    },
    {
      "epoch": 9.4542046063203,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.995,
      "step": 39715
    },
    {
      "epoch": 9.45444265904898,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 39716
    },
    {
      "epoch": 9.454680711777659,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0571,
      "step": 39717
    },
    {
      "epoch": 9.454918764506338,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.2457,
      "step": 39718
    },
    {
      "epoch": 9.455156817235018,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9373,
      "step": 39719
    },
    {
      "epoch": 9.455394869963698,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 39720
    },
    {
      "epoch": 9.455632922692377,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1054,
      "step": 39721
    },
    {
      "epoch": 9.455870975421055,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1164,
      "step": 39722
    },
    {
      "epoch": 9.456109028149735,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2399,
      "step": 39723
    },
    {
      "epoch": 9.456347080878414,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9011,
      "step": 39724
    },
    {
      "epoch": 9.456585133607094,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 39725
    },
    {
      "epoch": 9.456823186335773,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1549,
      "step": 39726
    },
    {
      "epoch": 9.457061239064453,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.871,
      "step": 39727
    },
    {
      "epoch": 9.457299291793133,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0284,
      "step": 39728
    },
    {
      "epoch": 9.457537344521812,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9317,
      "step": 39729
    },
    {
      "epoch": 9.45777539725049,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1301,
      "step": 39730
    },
    {
      "epoch": 9.45801344997917,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.7839,
      "step": 39731
    },
    {
      "epoch": 9.45825150270785,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1823,
      "step": 39732
    },
    {
      "epoch": 9.458489555436529,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.941,
      "step": 39733
    },
    {
      "epoch": 9.458727608165209,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.0141,
      "step": 39734
    },
    {
      "epoch": 9.458965660893888,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.7892,
      "step": 39735
    },
    {
      "epoch": 9.459203713622568,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 39736
    },
    {
      "epoch": 9.459441766351247,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3255,
      "step": 39737
    },
    {
      "epoch": 9.459679819079927,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2186,
      "step": 39738
    },
    {
      "epoch": 9.459917871808605,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9408,
      "step": 39739
    },
    {
      "epoch": 9.460155924537284,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 39740
    },
    {
      "epoch": 9.460393977265964,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1855,
      "step": 39741
    },
    {
      "epoch": 9.460632029994644,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0282,
      "step": 39742
    },
    {
      "epoch": 9.460870082723323,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1845,
      "step": 39743
    },
    {
      "epoch": 9.461108135452003,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7952,
      "step": 39744
    },
    {
      "epoch": 9.461346188180682,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0403,
      "step": 39745
    },
    {
      "epoch": 9.461584240909362,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.8467,
      "step": 39746
    },
    {
      "epoch": 9.461822293638042,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1113,
      "step": 39747
    },
    {
      "epoch": 9.46206034636672,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.7211,
      "step": 39748
    },
    {
      "epoch": 9.462298399095399,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 39749
    },
    {
      "epoch": 9.462536451824079,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8383,
      "step": 39750
    },
    {
      "epoch": 9.462774504552758,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9924,
      "step": 39751
    },
    {
      "epoch": 9.463012557281438,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1504,
      "step": 39752
    },
    {
      "epoch": 9.463250610010117,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9845,
      "step": 39753
    },
    {
      "epoch": 9.463488662738797,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.28,
      "step": 39754
    },
    {
      "epoch": 9.463726715467477,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9646,
      "step": 39755
    },
    {
      "epoch": 9.463964768196156,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7486,
      "step": 39756
    },
    {
      "epoch": 9.464202820924834,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1258,
      "step": 39757
    },
    {
      "epoch": 9.464440873653514,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 39758
    },
    {
      "epoch": 9.464678926382193,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8238,
      "step": 39759
    },
    {
      "epoch": 9.464916979110873,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8471,
      "step": 39760
    },
    {
      "epoch": 9.465155031839553,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9333,
      "step": 39761
    },
    {
      "epoch": 9.465393084568232,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 39762
    },
    {
      "epoch": 9.465631137296912,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2279,
      "step": 39763
    },
    {
      "epoch": 9.465869190025591,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.5511,
      "step": 39764
    },
    {
      "epoch": 9.46610724275427,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9382,
      "step": 39765
    },
    {
      "epoch": 9.466345295482949,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8036,
      "step": 39766
    },
    {
      "epoch": 9.466583348211628,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9201,
      "step": 39767
    },
    {
      "epoch": 9.466821400940308,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9968,
      "step": 39768
    },
    {
      "epoch": 9.467059453668988,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2687,
      "step": 39769
    },
    {
      "epoch": 9.467297506397667,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7492,
      "step": 39770
    },
    {
      "epoch": 9.467535559126347,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 39771
    },
    {
      "epoch": 9.467773611855026,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 39772
    },
    {
      "epoch": 9.468011664583706,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8241,
      "step": 39773
    },
    {
      "epoch": 9.468249717312384,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0954,
      "step": 39774
    },
    {
      "epoch": 9.468487770041063,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9673,
      "step": 39775
    },
    {
      "epoch": 9.468725822769743,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9748,
      "step": 39776
    },
    {
      "epoch": 9.468963875498423,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0059,
      "step": 39777
    },
    {
      "epoch": 9.469201928227102,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 39778
    },
    {
      "epoch": 9.469439980955782,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0716,
      "step": 39779
    },
    {
      "epoch": 9.469678033684461,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0489,
      "step": 39780
    },
    {
      "epoch": 9.469916086413141,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1169,
      "step": 39781
    },
    {
      "epoch": 9.47015413914182,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9294,
      "step": 39782
    },
    {
      "epoch": 9.470392191870499,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.16,
      "step": 39783
    },
    {
      "epoch": 9.470630244599178,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9953,
      "step": 39784
    },
    {
      "epoch": 9.470868297327858,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7871,
      "step": 39785
    },
    {
      "epoch": 9.471106350056537,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8919,
      "step": 39786
    },
    {
      "epoch": 9.471344402785217,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9336,
      "step": 39787
    },
    {
      "epoch": 9.471582455513897,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 39788
    },
    {
      "epoch": 9.471820508242576,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1079,
      "step": 39789
    },
    {
      "epoch": 9.472058560971256,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.994,
      "step": 39790
    },
    {
      "epoch": 9.472296613699935,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 39791
    },
    {
      "epoch": 9.472534666428613,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1099,
      "step": 39792
    },
    {
      "epoch": 9.472772719157293,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1139,
      "step": 39793
    },
    {
      "epoch": 9.473010771885972,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9173,
      "step": 39794
    },
    {
      "epoch": 9.473248824614652,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 39795
    },
    {
      "epoch": 9.473486877343332,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.8953,
      "step": 39796
    },
    {
      "epoch": 9.473724930072011,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8572,
      "step": 39797
    },
    {
      "epoch": 9.47396298280069,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.7868,
      "step": 39798
    },
    {
      "epoch": 9.47420103552937,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.808,
      "step": 39799
    },
    {
      "epoch": 9.474439088258048,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8628,
      "step": 39800
    },
    {
      "epoch": 9.474439088258048,
      "eval_runtime": 0.5288,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 39800
    },
    {
      "epoch": 9.474677140986728,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0534,
      "step": 39801
    },
    {
      "epoch": 9.474915193715407,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 1.9154,
      "step": 39802
    },
    {
      "epoch": 9.475153246444087,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2411,
      "step": 39803
    },
    {
      "epoch": 9.475391299172767,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0999,
      "step": 39804
    },
    {
      "epoch": 9.475629351901446,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0194,
      "step": 39805
    },
    {
      "epoch": 9.475867404630126,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9639,
      "step": 39806
    },
    {
      "epoch": 9.476105457358805,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0057,
      "step": 39807
    },
    {
      "epoch": 9.476343510087485,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9178,
      "step": 39808
    },
    {
      "epoch": 9.476581562816163,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.3247,
      "step": 39809
    },
    {
      "epoch": 9.476819615544843,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0664,
      "step": 39810
    },
    {
      "epoch": 9.477057668273522,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9696,
      "step": 39811
    },
    {
      "epoch": 9.477295721002202,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 39812
    },
    {
      "epoch": 9.477533773730881,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1372,
      "step": 39813
    },
    {
      "epoch": 9.477771826459561,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.797,
      "step": 39814
    },
    {
      "epoch": 9.47800987918824,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1147,
      "step": 39815
    },
    {
      "epoch": 9.47824793191692,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2363,
      "step": 39816
    },
    {
      "epoch": 9.4784859846456,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 1.914,
      "step": 39817
    },
    {
      "epoch": 9.478724037374278,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9117,
      "step": 39818
    },
    {
      "epoch": 9.478962090102957,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1165,
      "step": 39819
    },
    {
      "epoch": 9.479200142831637,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9873,
      "step": 39820
    },
    {
      "epoch": 9.479438195560316,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.3457,
      "step": 39821
    },
    {
      "epoch": 9.479676248288996,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.0714,
      "step": 39822
    },
    {
      "epoch": 9.479914301017676,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9091,
      "step": 39823
    },
    {
      "epoch": 9.480152353746355,
      "grad_norm": 7.125,
      "learning_rate": 3e-05,
      "loss": 2.1121,
      "step": 39824
    },
    {
      "epoch": 9.480390406475035,
      "grad_norm": 7.4375,
      "learning_rate": 3e-05,
      "loss": 2.0527,
      "step": 39825
    },
    {
      "epoch": 9.480628459203714,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0209,
      "step": 39826
    },
    {
      "epoch": 9.480866511932392,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.0267,
      "step": 39827
    },
    {
      "epoch": 9.481104564661072,
      "grad_norm": 7.96875,
      "learning_rate": 3e-05,
      "loss": 2.2068,
      "step": 39828
    },
    {
      "epoch": 9.481342617389751,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 39829
    },
    {
      "epoch": 9.481580670118431,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0537,
      "step": 39830
    },
    {
      "epoch": 9.48181872284711,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.4033,
      "step": 39831
    },
    {
      "epoch": 9.48205677557579,
      "grad_norm": 6.3125,
      "learning_rate": 3e-05,
      "loss": 2.3329,
      "step": 39832
    },
    {
      "epoch": 9.48229482830447,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0782,
      "step": 39833
    },
    {
      "epoch": 9.48253288103315,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 39834
    },
    {
      "epoch": 9.482770933761827,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.2927,
      "step": 39835
    },
    {
      "epoch": 9.483008986490507,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.2918,
      "step": 39836
    },
    {
      "epoch": 9.483247039219187,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9851,
      "step": 39837
    },
    {
      "epoch": 9.483485091947866,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0278,
      "step": 39838
    },
    {
      "epoch": 9.483723144676546,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1597,
      "step": 39839
    },
    {
      "epoch": 9.483961197405225,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.938,
      "step": 39840
    },
    {
      "epoch": 9.484199250133905,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9927,
      "step": 39841
    },
    {
      "epoch": 9.484437302862585,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9351,
      "step": 39842
    },
    {
      "epoch": 9.484675355591264,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0585,
      "step": 39843
    },
    {
      "epoch": 9.484913408319942,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2925,
      "step": 39844
    },
    {
      "epoch": 9.485151461048622,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0234,
      "step": 39845
    },
    {
      "epoch": 9.485389513777301,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8942,
      "step": 39846
    },
    {
      "epoch": 9.48562756650598,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.4057,
      "step": 39847
    },
    {
      "epoch": 9.48586561923466,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9015,
      "step": 39848
    },
    {
      "epoch": 9.48610367196334,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8425,
      "step": 39849
    },
    {
      "epoch": 9.48634172469202,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9106,
      "step": 39850
    },
    {
      "epoch": 9.4865797774207,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9328,
      "step": 39851
    },
    {
      "epoch": 9.486817830149379,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2137,
      "step": 39852
    },
    {
      "epoch": 9.487055882878057,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0261,
      "step": 39853
    },
    {
      "epoch": 9.487293935606736,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.205,
      "step": 39854
    },
    {
      "epoch": 9.487531988335416,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.971,
      "step": 39855
    },
    {
      "epoch": 9.487770041064095,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7206,
      "step": 39856
    },
    {
      "epoch": 9.488008093792775,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 39857
    },
    {
      "epoch": 9.488246146521455,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 39858
    },
    {
      "epoch": 9.488484199250134,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0009,
      "step": 39859
    },
    {
      "epoch": 9.488722251978814,
      "grad_norm": 2.796875,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 39860
    },
    {
      "epoch": 9.488960304707494,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9005,
      "step": 39861
    },
    {
      "epoch": 9.489198357436171,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 39862
    },
    {
      "epoch": 9.489436410164851,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0889,
      "step": 39863
    },
    {
      "epoch": 9.48967446289353,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0785,
      "step": 39864
    },
    {
      "epoch": 9.48991251562221,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 39865
    },
    {
      "epoch": 9.49015056835089,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0599,
      "step": 39866
    },
    {
      "epoch": 9.49038862107957,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1464,
      "step": 39867
    },
    {
      "epoch": 9.490626673808249,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1223,
      "step": 39868
    },
    {
      "epoch": 9.490864726536929,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.936,
      "step": 39869
    },
    {
      "epoch": 9.491102779265608,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9263,
      "step": 39870
    },
    {
      "epoch": 9.491340831994286,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.3242,
      "step": 39871
    },
    {
      "epoch": 9.491578884722966,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0323,
      "step": 39872
    },
    {
      "epoch": 9.491816937451645,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1974,
      "step": 39873
    },
    {
      "epoch": 9.492054990180325,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9923,
      "step": 39874
    },
    {
      "epoch": 9.492293042909004,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9286,
      "step": 39875
    },
    {
      "epoch": 9.492531095637684,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 39876
    },
    {
      "epoch": 9.492769148366364,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.158,
      "step": 39877
    },
    {
      "epoch": 9.493007201095043,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.3885,
      "step": 39878
    },
    {
      "epoch": 9.493245253823723,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.065,
      "step": 39879
    },
    {
      "epoch": 9.4934833065524,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7921,
      "step": 39880
    },
    {
      "epoch": 9.49372135928108,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9104,
      "step": 39881
    },
    {
      "epoch": 9.49395941200976,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.18,
      "step": 39882
    },
    {
      "epoch": 9.49419746473844,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.834,
      "step": 39883
    },
    {
      "epoch": 9.49443551746712,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 39884
    },
    {
      "epoch": 9.494673570195799,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9581,
      "step": 39885
    },
    {
      "epoch": 9.494911622924478,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 39886
    },
    {
      "epoch": 9.495149675653158,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0498,
      "step": 39887
    },
    {
      "epoch": 9.495387728381836,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1596,
      "step": 39888
    },
    {
      "epoch": 9.495625781110515,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9865,
      "step": 39889
    },
    {
      "epoch": 9.495863833839195,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1863,
      "step": 39890
    },
    {
      "epoch": 9.496101886567875,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 39891
    },
    {
      "epoch": 9.496339939296554,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0138,
      "step": 39892
    },
    {
      "epoch": 9.496577992025234,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9448,
      "step": 39893
    },
    {
      "epoch": 9.496816044753913,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2585,
      "step": 39894
    },
    {
      "epoch": 9.497054097482593,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9472,
      "step": 39895
    },
    {
      "epoch": 9.497292150211273,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1346,
      "step": 39896
    },
    {
      "epoch": 9.49753020293995,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4058,
      "step": 39897
    },
    {
      "epoch": 9.49776825566863,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0263,
      "step": 39898
    },
    {
      "epoch": 9.49800630839731,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.1063,
      "step": 39899
    },
    {
      "epoch": 9.49824436112599,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8546,
      "step": 39900
    },
    {
      "epoch": 9.49824436112599,
      "eval_runtime": 0.536,
      "eval_samples_per_second": 1.866,
      "eval_steps_per_second": 1.866,
      "step": 39900
    },
    {
      "epoch": 9.498482413854669,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8772,
      "step": 39901
    },
    {
      "epoch": 9.498720466583348,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2766,
      "step": 39902
    },
    {
      "epoch": 9.498958519312028,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5385,
      "step": 39903
    },
    {
      "epoch": 9.499196572040708,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 39904
    },
    {
      "epoch": 9.499434624769387,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1933,
      "step": 39905
    },
    {
      "epoch": 9.499672677498065,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0829,
      "step": 39906
    },
    {
      "epoch": 9.499910730226745,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.09,
      "step": 39907
    },
    {
      "epoch": 9.500148782955424,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 39908
    },
    {
      "epoch": 9.500386835684104,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.831,
      "step": 39909
    },
    {
      "epoch": 9.500624888412784,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.7797,
      "step": 39910
    },
    {
      "epoch": 9.500862941141463,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8833,
      "step": 39911
    },
    {
      "epoch": 9.501100993870143,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0789,
      "step": 39912
    },
    {
      "epoch": 9.501339046598822,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.2996,
      "step": 39913
    },
    {
      "epoch": 9.501577099327502,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8871,
      "step": 39914
    },
    {
      "epoch": 9.50181515205618,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2874,
      "step": 39915
    },
    {
      "epoch": 9.50205320478486,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9829,
      "step": 39916
    },
    {
      "epoch": 9.502291257513539,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8674,
      "step": 39917
    },
    {
      "epoch": 9.502529310242219,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0668,
      "step": 39918
    },
    {
      "epoch": 9.502767362970898,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 39919
    },
    {
      "epoch": 9.503005415699578,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3073,
      "step": 39920
    },
    {
      "epoch": 9.503243468428257,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2283,
      "step": 39921
    },
    {
      "epoch": 9.503481521156937,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 39922
    },
    {
      "epoch": 9.503719573885615,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 39923
    },
    {
      "epoch": 9.503957626614294,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8734,
      "step": 39924
    },
    {
      "epoch": 9.504195679342974,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9817,
      "step": 39925
    },
    {
      "epoch": 9.504433732071654,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2953,
      "step": 39926
    },
    {
      "epoch": 9.504671784800333,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9024,
      "step": 39927
    },
    {
      "epoch": 9.504909837529013,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1282,
      "step": 39928
    },
    {
      "epoch": 9.505147890257692,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 39929
    },
    {
      "epoch": 9.505385942986372,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.964,
      "step": 39930
    },
    {
      "epoch": 9.505623995715052,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.329,
      "step": 39931
    },
    {
      "epoch": 9.50586204844373,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1249,
      "step": 39932
    },
    {
      "epoch": 9.506100101172409,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1489,
      "step": 39933
    },
    {
      "epoch": 9.506338153901089,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 39934
    },
    {
      "epoch": 9.506576206629768,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 39935
    },
    {
      "epoch": 9.506814259358448,
      "grad_norm": 3.09375,
      "learning_rate": 3e-05,
      "loss": 2.1304,
      "step": 39936
    },
    {
      "epoch": 9.507052312087128,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.7487,
      "step": 39937
    },
    {
      "epoch": 9.507290364815807,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9383,
      "step": 39938
    },
    {
      "epoch": 9.507528417544487,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8746,
      "step": 39939
    },
    {
      "epoch": 9.507766470273165,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7742,
      "step": 39940
    },
    {
      "epoch": 9.508004523001844,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9908,
      "step": 39941
    },
    {
      "epoch": 9.508242575730524,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.3007,
      "step": 39942
    },
    {
      "epoch": 9.508480628459203,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 39943
    },
    {
      "epoch": 9.508718681187883,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 39944
    },
    {
      "epoch": 9.508956733916563,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8481,
      "step": 39945
    },
    {
      "epoch": 9.509194786645242,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.993,
      "step": 39946
    },
    {
      "epoch": 9.509432839373922,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.325,
      "step": 39947
    },
    {
      "epoch": 9.509670892102601,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 39948
    },
    {
      "epoch": 9.509908944831281,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.683,
      "step": 39949
    },
    {
      "epoch": 9.510146997559959,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.981,
      "step": 39950
    },
    {
      "epoch": 9.510385050288638,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 39951
    },
    {
      "epoch": 9.510623103017318,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.926,
      "step": 39952
    },
    {
      "epoch": 9.510861155745998,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 39953
    },
    {
      "epoch": 9.511099208474677,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9837,
      "step": 39954
    },
    {
      "epoch": 9.511337261203357,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2061,
      "step": 39955
    },
    {
      "epoch": 9.511575313932036,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 39956
    },
    {
      "epoch": 9.511813366660716,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.136,
      "step": 39957
    },
    {
      "epoch": 9.512051419389394,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2293,
      "step": 39958
    },
    {
      "epoch": 9.512289472118074,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.7914,
      "step": 39959
    },
    {
      "epoch": 9.512527524846753,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 39960
    },
    {
      "epoch": 9.512765577575433,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8583,
      "step": 39961
    },
    {
      "epoch": 9.513003630304112,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 39962
    },
    {
      "epoch": 9.513241683032792,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1257,
      "step": 39963
    },
    {
      "epoch": 9.513479735761472,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 39964
    },
    {
      "epoch": 9.513717788490151,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0871,
      "step": 39965
    },
    {
      "epoch": 9.51395584121883,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 39966
    },
    {
      "epoch": 9.514193893947509,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2275,
      "step": 39967
    },
    {
      "epoch": 9.514431946676188,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1194,
      "step": 39968
    },
    {
      "epoch": 9.514669999404868,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1816,
      "step": 39969
    },
    {
      "epoch": 9.514908052133547,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8769,
      "step": 39970
    },
    {
      "epoch": 9.515146104862227,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 39971
    },
    {
      "epoch": 9.515384157590907,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1585,
      "step": 39972
    },
    {
      "epoch": 9.515622210319586,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 39973
    },
    {
      "epoch": 9.515860263048266,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9312,
      "step": 39974
    },
    {
      "epoch": 9.516098315776944,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1532,
      "step": 39975
    },
    {
      "epoch": 9.516336368505623,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 39976
    },
    {
      "epoch": 9.516574421234303,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0442,
      "step": 39977
    },
    {
      "epoch": 9.516812473962982,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 39978
    },
    {
      "epoch": 9.517050526691662,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 39979
    },
    {
      "epoch": 9.517288579420342,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9177,
      "step": 39980
    },
    {
      "epoch": 9.517526632149021,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7268,
      "step": 39981
    },
    {
      "epoch": 9.5177646848777,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.8479,
      "step": 39982
    },
    {
      "epoch": 9.51800273760638,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0112,
      "step": 39983
    },
    {
      "epoch": 9.51824079033506,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.2432,
      "step": 39984
    },
    {
      "epoch": 9.518478843063738,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7574,
      "step": 39985
    },
    {
      "epoch": 9.518716895792418,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9654,
      "step": 39986
    },
    {
      "epoch": 9.518954948521097,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.287,
      "step": 39987
    },
    {
      "epoch": 9.519193001249777,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 39988
    },
    {
      "epoch": 9.519431053978456,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0653,
      "step": 39989
    },
    {
      "epoch": 9.519669106707136,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2538,
      "step": 39990
    },
    {
      "epoch": 9.519907159435816,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 39991
    },
    {
      "epoch": 9.520145212164495,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 39992
    },
    {
      "epoch": 9.520383264893173,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 39993
    },
    {
      "epoch": 9.520621317621853,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 39994
    },
    {
      "epoch": 9.520859370350532,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 39995
    },
    {
      "epoch": 9.521097423079212,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7532,
      "step": 39996
    },
    {
      "epoch": 9.521335475807891,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8927,
      "step": 39997
    },
    {
      "epoch": 9.521573528536571,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1473,
      "step": 39998
    },
    {
      "epoch": 9.52181158126525,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9792,
      "step": 39999
    },
    {
      "epoch": 9.52204963399393,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 40000
    },
    {
      "epoch": 9.52204963399393,
      "eval_runtime": 0.529,
      "eval_samples_per_second": 1.89,
      "eval_steps_per_second": 1.89,
      "step": 40000
    },
    {
      "epoch": 9.52228768672261,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0445,
      "step": 40001
    },
    {
      "epoch": 9.522525739451288,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 40002
    },
    {
      "epoch": 9.522763792179967,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9401,
      "step": 40003
    },
    {
      "epoch": 9.523001844908647,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8132,
      "step": 40004
    },
    {
      "epoch": 9.523239897637326,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.206,
      "step": 40005
    },
    {
      "epoch": 9.523477950366006,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2264,
      "step": 40006
    },
    {
      "epoch": 9.523716003094686,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 40007
    },
    {
      "epoch": 9.523954055823365,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 40008
    },
    {
      "epoch": 9.524192108552045,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2221,
      "step": 40009
    },
    {
      "epoch": 9.524430161280725,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2352,
      "step": 40010
    },
    {
      "epoch": 9.524668214009402,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1371,
      "step": 40011
    },
    {
      "epoch": 9.524906266738082,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2329,
      "step": 40012
    },
    {
      "epoch": 9.525144319466762,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 40013
    },
    {
      "epoch": 9.525382372195441,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0228,
      "step": 40014
    },
    {
      "epoch": 9.52562042492412,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1599,
      "step": 40015
    },
    {
      "epoch": 9.5258584776528,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.7703,
      "step": 40016
    },
    {
      "epoch": 9.52609653038148,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0849,
      "step": 40017
    },
    {
      "epoch": 9.52633458311016,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.3335,
      "step": 40018
    },
    {
      "epoch": 9.52657263583884,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9547,
      "step": 40019
    },
    {
      "epoch": 9.526810688567517,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 40020
    },
    {
      "epoch": 9.527048741296197,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1665,
      "step": 40021
    },
    {
      "epoch": 9.527286794024876,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 40022
    },
    {
      "epoch": 9.527524846753556,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2818,
      "step": 40023
    },
    {
      "epoch": 9.527762899482235,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2389,
      "step": 40024
    },
    {
      "epoch": 9.528000952210915,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1707,
      "step": 40025
    },
    {
      "epoch": 9.528239004939595,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 40026
    },
    {
      "epoch": 9.528477057668274,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0118,
      "step": 40027
    },
    {
      "epoch": 9.528715110396952,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0031,
      "step": 40028
    },
    {
      "epoch": 9.528953163125632,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 40029
    },
    {
      "epoch": 9.529191215854311,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9835,
      "step": 40030
    },
    {
      "epoch": 9.52942926858299,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0535,
      "step": 40031
    },
    {
      "epoch": 9.52966732131167,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 40032
    },
    {
      "epoch": 9.52990537404035,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.4385,
      "step": 40033
    },
    {
      "epoch": 9.53014342676903,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0437,
      "step": 40034
    },
    {
      "epoch": 9.53038147949771,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0641,
      "step": 40035
    },
    {
      "epoch": 9.530619532226389,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8638,
      "step": 40036
    },
    {
      "epoch": 9.530857584955067,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0061,
      "step": 40037
    },
    {
      "epoch": 9.531095637683746,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 40038
    },
    {
      "epoch": 9.531333690412426,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1579,
      "step": 40039
    },
    {
      "epoch": 9.531571743141106,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1043,
      "step": 40040
    },
    {
      "epoch": 9.531809795869785,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.8077,
      "step": 40041
    },
    {
      "epoch": 9.532047848598465,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3546,
      "step": 40042
    },
    {
      "epoch": 9.532285901327144,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 40043
    },
    {
      "epoch": 9.532523954055824,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.8222,
      "step": 40044
    },
    {
      "epoch": 9.532762006784504,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8209,
      "step": 40045
    },
    {
      "epoch": 9.533000059513181,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 40046
    },
    {
      "epoch": 9.533238112241861,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8104,
      "step": 40047
    },
    {
      "epoch": 9.53347616497054,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8203,
      "step": 40048
    },
    {
      "epoch": 9.53371421769922,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.9514,
      "step": 40049
    },
    {
      "epoch": 9.5339522704279,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1745,
      "step": 40050
    },
    {
      "epoch": 9.53419032315658,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9764,
      "step": 40051
    },
    {
      "epoch": 9.534428375885259,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.9974,
      "step": 40052
    },
    {
      "epoch": 9.534666428613939,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 40053
    },
    {
      "epoch": 9.534904481342618,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0273,
      "step": 40054
    },
    {
      "epoch": 9.535142534071296,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.9929,
      "step": 40055
    },
    {
      "epoch": 9.535380586799976,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.095,
      "step": 40056
    },
    {
      "epoch": 9.535618639528655,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.6954,
      "step": 40057
    },
    {
      "epoch": 9.535856692257335,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.875,
      "step": 40058
    },
    {
      "epoch": 9.536094744986015,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.175,
      "step": 40059
    },
    {
      "epoch": 9.536332797714694,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 40060
    },
    {
      "epoch": 9.536570850443374,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 1.803,
      "step": 40061
    },
    {
      "epoch": 9.536808903172053,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.147,
      "step": 40062
    },
    {
      "epoch": 9.537046955900731,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2815,
      "step": 40063
    },
    {
      "epoch": 9.53728500862941,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9814,
      "step": 40064
    },
    {
      "epoch": 9.53752306135809,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.905,
      "step": 40065
    },
    {
      "epoch": 9.53776111408677,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2159,
      "step": 40066
    },
    {
      "epoch": 9.53799916681545,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 40067
    },
    {
      "epoch": 9.53823721954413,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 40068
    },
    {
      "epoch": 9.538475272272809,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3078,
      "step": 40069
    },
    {
      "epoch": 9.538713325001488,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.1525,
      "step": 40070
    },
    {
      "epoch": 9.538951377730168,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1047,
      "step": 40071
    },
    {
      "epoch": 9.539189430458846,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.5936,
      "step": 40072
    },
    {
      "epoch": 9.539427483187525,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8534,
      "step": 40073
    },
    {
      "epoch": 9.539665535916205,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9025,
      "step": 40074
    },
    {
      "epoch": 9.539903588644885,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.3095,
      "step": 40075
    },
    {
      "epoch": 9.540141641373564,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.2427,
      "step": 40076
    },
    {
      "epoch": 9.540379694102244,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 40077
    },
    {
      "epoch": 9.540617746830923,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 40078
    },
    {
      "epoch": 9.540855799559603,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9676,
      "step": 40079
    },
    {
      "epoch": 9.541093852288283,
      "grad_norm": 6.15625,
      "learning_rate": 3e-05,
      "loss": 2.252,
      "step": 40080
    },
    {
      "epoch": 9.54133190501696,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9823,
      "step": 40081
    },
    {
      "epoch": 9.54156995774564,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.251,
      "step": 40082
    },
    {
      "epoch": 9.54180801047432,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0655,
      "step": 40083
    },
    {
      "epoch": 9.542046063203,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 40084
    },
    {
      "epoch": 9.542284115931679,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1027,
      "step": 40085
    },
    {
      "epoch": 9.542522168660359,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.87,
      "step": 40086
    },
    {
      "epoch": 9.542760221389038,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7389,
      "step": 40087
    },
    {
      "epoch": 9.542998274117718,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8446,
      "step": 40088
    },
    {
      "epoch": 9.543236326846397,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.781,
      "step": 40089
    },
    {
      "epoch": 9.543474379575075,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.04,
      "step": 40090
    },
    {
      "epoch": 9.543712432303755,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0619,
      "step": 40091
    },
    {
      "epoch": 9.543950485032434,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.908,
      "step": 40092
    },
    {
      "epoch": 9.544188537761114,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8343,
      "step": 40093
    },
    {
      "epoch": 9.544426590489794,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0493,
      "step": 40094
    },
    {
      "epoch": 9.544664643218473,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.815,
      "step": 40095
    },
    {
      "epoch": 9.544902695947153,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0319,
      "step": 40096
    },
    {
      "epoch": 9.545140748675832,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9826,
      "step": 40097
    },
    {
      "epoch": 9.54537880140451,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8854,
      "step": 40098
    },
    {
      "epoch": 9.54561685413319,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.1674,
      "step": 40099
    },
    {
      "epoch": 9.54585490686187,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.192,
      "step": 40100
    },
    {
      "epoch": 9.54585490686187,
      "eval_runtime": 0.52,
      "eval_samples_per_second": 1.923,
      "eval_steps_per_second": 1.923,
      "step": 40100
    },
    {
      "epoch": 9.546092959590549,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.7702,
      "step": 40101
    },
    {
      "epoch": 9.546331012319229,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0729,
      "step": 40102
    },
    {
      "epoch": 9.546569065047908,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.8662,
      "step": 40103
    },
    {
      "epoch": 9.546807117776588,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.961,
      "step": 40104
    },
    {
      "epoch": 9.547045170505267,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.96,
      "step": 40105
    },
    {
      "epoch": 9.547283223233947,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.8255,
      "step": 40106
    },
    {
      "epoch": 9.547521275962625,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9798,
      "step": 40107
    },
    {
      "epoch": 9.547759328691305,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 40108
    },
    {
      "epoch": 9.547997381419984,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0054,
      "step": 40109
    },
    {
      "epoch": 9.548235434148664,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0941,
      "step": 40110
    },
    {
      "epoch": 9.548473486877343,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 40111
    },
    {
      "epoch": 9.548711539606023,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1634,
      "step": 40112
    },
    {
      "epoch": 9.548949592334703,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8924,
      "step": 40113
    },
    {
      "epoch": 9.549187645063382,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.3104,
      "step": 40114
    },
    {
      "epoch": 9.549425697792062,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.008,
      "step": 40115
    },
    {
      "epoch": 9.54966375052074,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8274,
      "step": 40116
    },
    {
      "epoch": 9.54990180324942,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0543,
      "step": 40117
    },
    {
      "epoch": 9.550139855978099,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 40118
    },
    {
      "epoch": 9.550377908706778,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0684,
      "step": 40119
    },
    {
      "epoch": 9.550615961435458,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 40120
    },
    {
      "epoch": 9.550854014164138,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 40121
    },
    {
      "epoch": 9.551092066892817,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9035,
      "step": 40122
    },
    {
      "epoch": 9.551330119621497,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9699,
      "step": 40123
    },
    {
      "epoch": 9.551568172350176,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2039,
      "step": 40124
    },
    {
      "epoch": 9.551806225078854,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2023,
      "step": 40125
    },
    {
      "epoch": 9.552044277807534,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9828,
      "step": 40126
    },
    {
      "epoch": 9.552282330536213,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 40127
    },
    {
      "epoch": 9.552520383264893,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7715,
      "step": 40128
    },
    {
      "epoch": 9.552758435993573,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 2.0924,
      "step": 40129
    },
    {
      "epoch": 9.552996488722252,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.1196,
      "step": 40130
    },
    {
      "epoch": 9.553234541450932,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9043,
      "step": 40131
    },
    {
      "epoch": 9.553472594179611,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7754,
      "step": 40132
    },
    {
      "epoch": 9.55371064690829,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 40133
    },
    {
      "epoch": 9.553948699636969,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2769,
      "step": 40134
    },
    {
      "epoch": 9.554186752365649,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.9815,
      "step": 40135
    },
    {
      "epoch": 9.554424805094328,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1915,
      "step": 40136
    },
    {
      "epoch": 9.554662857823008,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.3953,
      "step": 40137
    },
    {
      "epoch": 9.554900910551687,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2661,
      "step": 40138
    },
    {
      "epoch": 9.555138963280367,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0279,
      "step": 40139
    },
    {
      "epoch": 9.555377016009047,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0801,
      "step": 40140
    },
    {
      "epoch": 9.555615068737726,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0788,
      "step": 40141
    },
    {
      "epoch": 9.555853121466404,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9102,
      "step": 40142
    },
    {
      "epoch": 9.556091174195084,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9506,
      "step": 40143
    },
    {
      "epoch": 9.556329226923763,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0944,
      "step": 40144
    },
    {
      "epoch": 9.556567279652443,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0036,
      "step": 40145
    },
    {
      "epoch": 9.556805332381122,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 40146
    },
    {
      "epoch": 9.557043385109802,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9536,
      "step": 40147
    },
    {
      "epoch": 9.557281437838482,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8674,
      "step": 40148
    },
    {
      "epoch": 9.557519490567161,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.6622,
      "step": 40149
    },
    {
      "epoch": 9.55775754329584,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 40150
    },
    {
      "epoch": 9.557995596024519,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2213,
      "step": 40151
    },
    {
      "epoch": 9.558233648753198,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3139,
      "step": 40152
    },
    {
      "epoch": 9.558471701481878,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 1.7996,
      "step": 40153
    },
    {
      "epoch": 9.558709754210557,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.6939,
      "step": 40154
    },
    {
      "epoch": 9.558947806939237,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0336,
      "step": 40155
    },
    {
      "epoch": 9.559185859667917,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 40156
    },
    {
      "epoch": 9.559423912396596,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 40157
    },
    {
      "epoch": 9.559661965125276,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2362,
      "step": 40158
    },
    {
      "epoch": 9.559900017853955,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2182,
      "step": 40159
    },
    {
      "epoch": 9.560138070582633,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8843,
      "step": 40160
    },
    {
      "epoch": 9.560376123311313,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9795,
      "step": 40161
    },
    {
      "epoch": 9.560614176039993,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 40162
    },
    {
      "epoch": 9.560852228768672,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 40163
    },
    {
      "epoch": 9.561090281497352,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0583,
      "step": 40164
    },
    {
      "epoch": 9.561328334226031,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2296,
      "step": 40165
    },
    {
      "epoch": 9.561566386954711,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 40166
    },
    {
      "epoch": 9.56180443968339,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 40167
    },
    {
      "epoch": 9.562042492412068,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8409,
      "step": 40168
    },
    {
      "epoch": 9.562280545140748,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7613,
      "step": 40169
    },
    {
      "epoch": 9.562518597869428,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 40170
    },
    {
      "epoch": 9.562756650598107,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9839,
      "step": 40171
    },
    {
      "epoch": 9.562994703326787,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.3539,
      "step": 40172
    },
    {
      "epoch": 9.563232756055466,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2966,
      "step": 40173
    },
    {
      "epoch": 9.563470808784146,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0377,
      "step": 40174
    },
    {
      "epoch": 9.563708861512826,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8125,
      "step": 40175
    },
    {
      "epoch": 9.563946914241505,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 40176
    },
    {
      "epoch": 9.564184966970185,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9184,
      "step": 40177
    },
    {
      "epoch": 9.564423019698863,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1695,
      "step": 40178
    },
    {
      "epoch": 9.564661072427542,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0129,
      "step": 40179
    },
    {
      "epoch": 9.564899125156222,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9858,
      "step": 40180
    },
    {
      "epoch": 9.565137177884901,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8891,
      "step": 40181
    },
    {
      "epoch": 9.565375230613581,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 40182
    },
    {
      "epoch": 9.56561328334226,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 40183
    },
    {
      "epoch": 9.56585133607094,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0487,
      "step": 40184
    },
    {
      "epoch": 9.56608938879962,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1387,
      "step": 40185
    },
    {
      "epoch": 9.566327441528298,
      "grad_norm": 6.375,
      "learning_rate": 3e-05,
      "loss": 2.5025,
      "step": 40186
    },
    {
      "epoch": 9.566565494256977,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9038,
      "step": 40187
    },
    {
      "epoch": 9.566803546985657,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 40188
    },
    {
      "epoch": 9.567041599714337,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9375,
      "step": 40189
    },
    {
      "epoch": 9.567279652443016,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9415,
      "step": 40190
    },
    {
      "epoch": 9.567517705171696,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 40191
    },
    {
      "epoch": 9.567755757900375,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1032,
      "step": 40192
    },
    {
      "epoch": 9.567993810629055,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0969,
      "step": 40193
    },
    {
      "epoch": 9.568231863357735,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.113,
      "step": 40194
    },
    {
      "epoch": 9.568469916086412,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0401,
      "step": 40195
    },
    {
      "epoch": 9.568707968815092,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9036,
      "step": 40196
    },
    {
      "epoch": 9.568946021543772,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1751,
      "step": 40197
    },
    {
      "epoch": 9.569184074272451,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1237,
      "step": 40198
    },
    {
      "epoch": 9.56942212700113,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9156,
      "step": 40199
    },
    {
      "epoch": 9.56966017972981,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0256,
      "step": 40200
    },
    {
      "epoch": 9.56966017972981,
      "eval_runtime": 0.5267,
      "eval_samples_per_second": 1.899,
      "eval_steps_per_second": 1.899,
      "step": 40200
    },
    {
      "epoch": 9.56989823245849,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2651,
      "step": 40201
    },
    {
      "epoch": 9.57013628518717,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1025,
      "step": 40202
    },
    {
      "epoch": 9.570374337915847,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2883,
      "step": 40203
    },
    {
      "epoch": 9.570612390644527,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9918,
      "step": 40204
    },
    {
      "epoch": 9.570850443373207,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9724,
      "step": 40205
    },
    {
      "epoch": 9.571088496101886,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0631,
      "step": 40206
    },
    {
      "epoch": 9.571326548830566,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 40207
    },
    {
      "epoch": 9.571564601559245,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.0823,
      "step": 40208
    },
    {
      "epoch": 9.571802654287925,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7603,
      "step": 40209
    },
    {
      "epoch": 9.572040707016605,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1888,
      "step": 40210
    },
    {
      "epoch": 9.572278759745284,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 40211
    },
    {
      "epoch": 9.572516812473964,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.3662,
      "step": 40212
    },
    {
      "epoch": 9.572754865202642,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 40213
    },
    {
      "epoch": 9.572992917931321,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.9444,
      "step": 40214
    },
    {
      "epoch": 9.573230970660001,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2009,
      "step": 40215
    },
    {
      "epoch": 9.57346902338868,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.152,
      "step": 40216
    },
    {
      "epoch": 9.57370707611736,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9682,
      "step": 40217
    },
    {
      "epoch": 9.57394512884604,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.3335,
      "step": 40218
    },
    {
      "epoch": 9.57418318157472,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.6004,
      "step": 40219
    },
    {
      "epoch": 9.574421234303399,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8931,
      "step": 40220
    },
    {
      "epoch": 9.574659287032077,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0809,
      "step": 40221
    },
    {
      "epoch": 9.574897339760756,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8551,
      "step": 40222
    },
    {
      "epoch": 9.575135392489436,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.154,
      "step": 40223
    },
    {
      "epoch": 9.575373445218116,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1601,
      "step": 40224
    },
    {
      "epoch": 9.575611497946795,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2013,
      "step": 40225
    },
    {
      "epoch": 9.575849550675475,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.3785,
      "step": 40226
    },
    {
      "epoch": 9.576087603404154,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8879,
      "step": 40227
    },
    {
      "epoch": 9.576325656132834,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1167,
      "step": 40228
    },
    {
      "epoch": 9.576563708861514,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.1166,
      "step": 40229
    },
    {
      "epoch": 9.576801761590191,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.5907,
      "step": 40230
    },
    {
      "epoch": 9.577039814318871,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7293,
      "step": 40231
    },
    {
      "epoch": 9.57727786704755,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 40232
    },
    {
      "epoch": 9.57751591977623,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9436,
      "step": 40233
    },
    {
      "epoch": 9.57775397250491,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2251,
      "step": 40234
    },
    {
      "epoch": 9.57799202523359,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 40235
    },
    {
      "epoch": 9.578230077962269,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.132,
      "step": 40236
    },
    {
      "epoch": 9.578468130690949,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.382,
      "step": 40237
    },
    {
      "epoch": 9.578706183419627,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 40238
    },
    {
      "epoch": 9.578944236148306,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 40239
    },
    {
      "epoch": 9.579182288876986,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9708,
      "step": 40240
    },
    {
      "epoch": 9.579420341605665,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 40241
    },
    {
      "epoch": 9.579658394334345,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 40242
    },
    {
      "epoch": 9.579896447063025,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 40243
    },
    {
      "epoch": 9.580134499791704,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.9716,
      "step": 40244
    },
    {
      "epoch": 9.580372552520384,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 40245
    },
    {
      "epoch": 9.580610605249063,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0959,
      "step": 40246
    },
    {
      "epoch": 9.580848657977743,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 40247
    },
    {
      "epoch": 9.58108671070642,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1105,
      "step": 40248
    },
    {
      "epoch": 9.5813247634351,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 40249
    },
    {
      "epoch": 9.58156281616378,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1339,
      "step": 40250
    },
    {
      "epoch": 9.58180086889246,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8258,
      "step": 40251
    },
    {
      "epoch": 9.58203892162114,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 40252
    },
    {
      "epoch": 9.582276974349819,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1499,
      "step": 40253
    },
    {
      "epoch": 9.582515027078498,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9641,
      "step": 40254
    },
    {
      "epoch": 9.582753079807178,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1272,
      "step": 40255
    },
    {
      "epoch": 9.582991132535856,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8453,
      "step": 40256
    },
    {
      "epoch": 9.583229185264535,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.0508,
      "step": 40257
    },
    {
      "epoch": 9.583467237993215,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0258,
      "step": 40258
    },
    {
      "epoch": 9.583705290721895,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9489,
      "step": 40259
    },
    {
      "epoch": 9.583943343450574,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0187,
      "step": 40260
    },
    {
      "epoch": 9.584181396179254,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 40261
    },
    {
      "epoch": 9.584419448907934,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0108,
      "step": 40262
    },
    {
      "epoch": 9.584657501636613,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 40263
    },
    {
      "epoch": 9.584895554365293,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0554,
      "step": 40264
    },
    {
      "epoch": 9.58513360709397,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 40265
    },
    {
      "epoch": 9.58537165982265,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 40266
    },
    {
      "epoch": 9.58560971255133,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1803,
      "step": 40267
    },
    {
      "epoch": 9.58584776528001,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1662,
      "step": 40268
    },
    {
      "epoch": 9.586085818008689,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.836,
      "step": 40269
    },
    {
      "epoch": 9.586323870737369,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1856,
      "step": 40270
    },
    {
      "epoch": 9.586561923466048,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1518,
      "step": 40271
    },
    {
      "epoch": 9.586799976194728,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 40272
    },
    {
      "epoch": 9.587038028923406,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0752,
      "step": 40273
    },
    {
      "epoch": 9.587276081652085,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.118,
      "step": 40274
    },
    {
      "epoch": 9.587514134380765,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0796,
      "step": 40275
    },
    {
      "epoch": 9.587752187109444,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 40276
    },
    {
      "epoch": 9.587990239838124,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.0625,
      "step": 40277
    },
    {
      "epoch": 9.588228292566804,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.3163,
      "step": 40278
    },
    {
      "epoch": 9.588466345295483,
      "grad_norm": 5.65625,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 40279
    },
    {
      "epoch": 9.588704398024163,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1456,
      "step": 40280
    },
    {
      "epoch": 9.588942450752842,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 40281
    },
    {
      "epoch": 9.589180503481522,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2028,
      "step": 40282
    },
    {
      "epoch": 9.5894185562102,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 40283
    },
    {
      "epoch": 9.58965660893888,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9163,
      "step": 40284
    },
    {
      "epoch": 9.589894661667559,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9473,
      "step": 40285
    },
    {
      "epoch": 9.590132714396239,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.2053,
      "step": 40286
    },
    {
      "epoch": 9.590370767124918,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 40287
    },
    {
      "epoch": 9.590608819853598,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 40288
    },
    {
      "epoch": 9.590846872582278,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1982,
      "step": 40289
    },
    {
      "epoch": 9.591084925310957,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1294,
      "step": 40290
    },
    {
      "epoch": 9.591322978039635,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8655,
      "step": 40291
    },
    {
      "epoch": 9.591561030768315,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0196,
      "step": 40292
    },
    {
      "epoch": 9.591799083496994,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9323,
      "step": 40293
    },
    {
      "epoch": 9.592037136225674,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 40294
    },
    {
      "epoch": 9.592275188954353,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.4583,
      "step": 40295
    },
    {
      "epoch": 9.592513241683033,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 40296
    },
    {
      "epoch": 9.592751294411713,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7174,
      "step": 40297
    },
    {
      "epoch": 9.592989347140392,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 40298
    },
    {
      "epoch": 9.593227399869072,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9474,
      "step": 40299
    },
    {
      "epoch": 9.59346545259775,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9551,
      "step": 40300
    },
    {
      "epoch": 9.59346545259775,
      "eval_runtime": 0.5131,
      "eval_samples_per_second": 1.949,
      "eval_steps_per_second": 1.949,
      "step": 40300
    },
    {
      "epoch": 9.59370350532643,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 40301
    },
    {
      "epoch": 9.593941558055109,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2904,
      "step": 40302
    },
    {
      "epoch": 9.594179610783788,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2415,
      "step": 40303
    },
    {
      "epoch": 9.594417663512468,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 40304
    },
    {
      "epoch": 9.594655716241148,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.184,
      "step": 40305
    },
    {
      "epoch": 9.594893768969827,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1183,
      "step": 40306
    },
    {
      "epoch": 9.595131821698507,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0914,
      "step": 40307
    },
    {
      "epoch": 9.595369874427185,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.248,
      "step": 40308
    },
    {
      "epoch": 9.595607927155864,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 40309
    },
    {
      "epoch": 9.595845979884544,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0587,
      "step": 40310
    },
    {
      "epoch": 9.596084032613224,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8418,
      "step": 40311
    },
    {
      "epoch": 9.596322085341903,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0227,
      "step": 40312
    },
    {
      "epoch": 9.596560138070583,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 40313
    },
    {
      "epoch": 9.596798190799262,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9359,
      "step": 40314
    },
    {
      "epoch": 9.597036243527942,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 40315
    },
    {
      "epoch": 9.597274296256622,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 40316
    },
    {
      "epoch": 9.597512348985301,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2073,
      "step": 40317
    },
    {
      "epoch": 9.597750401713979,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1859,
      "step": 40318
    },
    {
      "epoch": 9.597988454442659,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0521,
      "step": 40319
    },
    {
      "epoch": 9.598226507171338,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0614,
      "step": 40320
    },
    {
      "epoch": 9.598464559900018,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9904,
      "step": 40321
    },
    {
      "epoch": 9.598702612628697,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.155,
      "step": 40322
    },
    {
      "epoch": 9.598940665357377,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0271,
      "step": 40323
    },
    {
      "epoch": 9.599178718086057,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2394,
      "step": 40324
    },
    {
      "epoch": 9.599416770814736,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8888,
      "step": 40325
    },
    {
      "epoch": 9.599654823543414,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9808,
      "step": 40326
    },
    {
      "epoch": 9.599892876272094,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8248,
      "step": 40327
    },
    {
      "epoch": 9.600130929000773,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 40328
    },
    {
      "epoch": 9.600368981729453,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.4263,
      "step": 40329
    },
    {
      "epoch": 9.600607034458132,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9311,
      "step": 40330
    },
    {
      "epoch": 9.600845087186812,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9293,
      "step": 40331
    },
    {
      "epoch": 9.601083139915492,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 40332
    },
    {
      "epoch": 9.601321192644171,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.2376,
      "step": 40333
    },
    {
      "epoch": 9.60155924537285,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0467,
      "step": 40334
    },
    {
      "epoch": 9.601797298101529,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0962,
      "step": 40335
    },
    {
      "epoch": 9.602035350830208,
      "grad_norm": 6.21875,
      "learning_rate": 3e-05,
      "loss": 2.0354,
      "step": 40336
    },
    {
      "epoch": 9.602273403558888,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 40337
    },
    {
      "epoch": 9.602511456287568,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0998,
      "step": 40338
    },
    {
      "epoch": 9.602749509016247,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.048,
      "step": 40339
    },
    {
      "epoch": 9.602987561744927,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1766,
      "step": 40340
    },
    {
      "epoch": 9.603225614473606,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0147,
      "step": 40341
    },
    {
      "epoch": 9.603463667202286,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.9693,
      "step": 40342
    },
    {
      "epoch": 9.603701719930964,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0183,
      "step": 40343
    },
    {
      "epoch": 9.603939772659643,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.897,
      "step": 40344
    },
    {
      "epoch": 9.604177825388323,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.9571,
      "step": 40345
    },
    {
      "epoch": 9.604415878117003,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.282,
      "step": 40346
    },
    {
      "epoch": 9.604653930845682,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8321,
      "step": 40347
    },
    {
      "epoch": 9.604891983574362,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.3372,
      "step": 40348
    },
    {
      "epoch": 9.605130036303041,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9029,
      "step": 40349
    },
    {
      "epoch": 9.605368089031721,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.8955,
      "step": 40350
    },
    {
      "epoch": 9.6056061417604,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0612,
      "step": 40351
    },
    {
      "epoch": 9.60584419448908,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.8477,
      "step": 40352
    },
    {
      "epoch": 9.606082247217758,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9095,
      "step": 40353
    },
    {
      "epoch": 9.606320299946438,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0251,
      "step": 40354
    },
    {
      "epoch": 9.606558352675117,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9686,
      "step": 40355
    },
    {
      "epoch": 9.606796405403797,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 40356
    },
    {
      "epoch": 9.607034458132476,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 40357
    },
    {
      "epoch": 9.607272510861156,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0566,
      "step": 40358
    },
    {
      "epoch": 9.607510563589836,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8733,
      "step": 40359
    },
    {
      "epoch": 9.607748616318515,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 40360
    },
    {
      "epoch": 9.607986669047193,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 40361
    },
    {
      "epoch": 9.608224721775873,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1308,
      "step": 40362
    },
    {
      "epoch": 9.608462774504552,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8812,
      "step": 40363
    },
    {
      "epoch": 9.608700827233232,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 40364
    },
    {
      "epoch": 9.608938879961912,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8298,
      "step": 40365
    },
    {
      "epoch": 9.609176932690591,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1094,
      "step": 40366
    },
    {
      "epoch": 9.60941498541927,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.1732,
      "step": 40367
    },
    {
      "epoch": 9.60965303814795,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 40368
    },
    {
      "epoch": 9.60989109087663,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 40369
    },
    {
      "epoch": 9.610129143605308,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0023,
      "step": 40370
    },
    {
      "epoch": 9.610367196333987,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 40371
    },
    {
      "epoch": 9.610605249062667,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.792,
      "step": 40372
    },
    {
      "epoch": 9.610843301791347,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.045,
      "step": 40373
    },
    {
      "epoch": 9.611081354520026,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.7843,
      "step": 40374
    },
    {
      "epoch": 9.611319407248706,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1984,
      "step": 40375
    },
    {
      "epoch": 9.611557459977385,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.1869,
      "step": 40376
    },
    {
      "epoch": 9.611795512706065,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9714,
      "step": 40377
    },
    {
      "epoch": 9.612033565434743,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1221,
      "step": 40378
    },
    {
      "epoch": 9.612271618163422,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9128,
      "step": 40379
    },
    {
      "epoch": 9.612509670892102,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0874,
      "step": 40380
    },
    {
      "epoch": 9.612747723620782,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 40381
    },
    {
      "epoch": 9.612985776349461,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8744,
      "step": 40382
    },
    {
      "epoch": 9.61322382907814,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 40383
    },
    {
      "epoch": 9.61346188180682,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.2072,
      "step": 40384
    },
    {
      "epoch": 9.6136999345355,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1382,
      "step": 40385
    },
    {
      "epoch": 9.61393798726418,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.9734,
      "step": 40386
    },
    {
      "epoch": 9.61417603999286,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1029,
      "step": 40387
    },
    {
      "epoch": 9.614414092721537,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 2.1252,
      "step": 40388
    },
    {
      "epoch": 9.614652145450217,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8695,
      "step": 40389
    },
    {
      "epoch": 9.614890198178896,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 2.224,
      "step": 40390
    },
    {
      "epoch": 9.615128250907576,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.4815,
      "step": 40391
    },
    {
      "epoch": 9.615366303636256,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0758,
      "step": 40392
    },
    {
      "epoch": 9.615604356364935,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2672,
      "step": 40393
    },
    {
      "epoch": 9.615842409093615,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 40394
    },
    {
      "epoch": 9.616080461822294,
      "grad_norm": 3.0,
      "learning_rate": 3e-05,
      "loss": 1.9576,
      "step": 40395
    },
    {
      "epoch": 9.616318514550972,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9197,
      "step": 40396
    },
    {
      "epoch": 9.616556567279652,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.2688,
      "step": 40397
    },
    {
      "epoch": 9.616794620008331,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.6529,
      "step": 40398
    },
    {
      "epoch": 9.617032672737011,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.3672,
      "step": 40399
    },
    {
      "epoch": 9.61727072546569,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7346,
      "step": 40400
    },
    {
      "epoch": 9.61727072546569,
      "eval_runtime": 0.5246,
      "eval_samples_per_second": 1.906,
      "eval_steps_per_second": 1.906,
      "step": 40400
    },
    {
      "epoch": 9.61750877819437,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0483,
      "step": 40401
    },
    {
      "epoch": 9.61774683092305,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1951,
      "step": 40402
    },
    {
      "epoch": 9.61798488365173,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9621,
      "step": 40403
    },
    {
      "epoch": 9.618222936380409,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 40404
    },
    {
      "epoch": 9.618460989109087,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0192,
      "step": 40405
    },
    {
      "epoch": 9.618699041837766,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.1006,
      "step": 40406
    },
    {
      "epoch": 9.618937094566446,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1715,
      "step": 40407
    },
    {
      "epoch": 9.619175147295126,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 40408
    },
    {
      "epoch": 9.619413200023805,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9047,
      "step": 40409
    },
    {
      "epoch": 9.619651252752485,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 40410
    },
    {
      "epoch": 9.619889305481164,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 2.0162,
      "step": 40411
    },
    {
      "epoch": 9.620127358209844,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 40412
    },
    {
      "epoch": 9.620365410938524,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9819,
      "step": 40413
    },
    {
      "epoch": 9.620603463667202,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 40414
    },
    {
      "epoch": 9.620841516395881,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 40415
    },
    {
      "epoch": 9.62107956912456,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1286,
      "step": 40416
    },
    {
      "epoch": 9.62131762185324,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.5236,
      "step": 40417
    },
    {
      "epoch": 9.62155567458192,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8516,
      "step": 40418
    },
    {
      "epoch": 9.6217937273106,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 2.0818,
      "step": 40419
    },
    {
      "epoch": 9.62203178003928,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1868,
      "step": 40420
    },
    {
      "epoch": 9.622269832767959,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9593,
      "step": 40421
    },
    {
      "epoch": 9.622507885496638,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9318,
      "step": 40422
    },
    {
      "epoch": 9.622745938225316,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9115,
      "step": 40423
    },
    {
      "epoch": 9.622983990953996,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1128,
      "step": 40424
    },
    {
      "epoch": 9.623222043682675,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1159,
      "step": 40425
    },
    {
      "epoch": 9.623460096411355,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 40426
    },
    {
      "epoch": 9.623698149140035,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0858,
      "step": 40427
    },
    {
      "epoch": 9.623936201868714,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0663,
      "step": 40428
    },
    {
      "epoch": 9.624174254597394,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8743,
      "step": 40429
    },
    {
      "epoch": 9.624412307326073,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 40430
    },
    {
      "epoch": 9.624650360054751,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0328,
      "step": 40431
    },
    {
      "epoch": 9.62488841278343,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9973,
      "step": 40432
    },
    {
      "epoch": 9.62512646551211,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 40433
    },
    {
      "epoch": 9.62536451824079,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2653,
      "step": 40434
    },
    {
      "epoch": 9.62560257096947,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 40435
    },
    {
      "epoch": 9.62584062369815,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0363,
      "step": 40436
    },
    {
      "epoch": 9.626078676426829,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.128,
      "step": 40437
    },
    {
      "epoch": 9.626316729155509,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0698,
      "step": 40438
    },
    {
      "epoch": 9.626554781884188,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0567,
      "step": 40439
    },
    {
      "epoch": 9.626792834612866,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 40440
    },
    {
      "epoch": 9.627030887341546,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8364,
      "step": 40441
    },
    {
      "epoch": 9.627268940070225,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.8004,
      "step": 40442
    },
    {
      "epoch": 9.627506992798905,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0945,
      "step": 40443
    },
    {
      "epoch": 9.627745045527584,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.7882,
      "step": 40444
    },
    {
      "epoch": 9.627983098256264,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1645,
      "step": 40445
    },
    {
      "epoch": 9.628221150984944,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.849,
      "step": 40446
    },
    {
      "epoch": 9.628459203713623,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.1655,
      "step": 40447
    },
    {
      "epoch": 9.628697256442303,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0821,
      "step": 40448
    },
    {
      "epoch": 9.62893530917098,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8562,
      "step": 40449
    },
    {
      "epoch": 9.62917336189966,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.553,
      "step": 40450
    },
    {
      "epoch": 9.62941141462834,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.887,
      "step": 40451
    },
    {
      "epoch": 9.62964946735702,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.9963,
      "step": 40452
    },
    {
      "epoch": 9.629887520085699,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.209,
      "step": 40453
    },
    {
      "epoch": 9.630125572814379,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9325,
      "step": 40454
    },
    {
      "epoch": 9.630363625543058,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8683,
      "step": 40455
    },
    {
      "epoch": 9.630601678271738,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 40456
    },
    {
      "epoch": 9.630839731000417,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9701,
      "step": 40457
    },
    {
      "epoch": 9.631077783729095,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8163,
      "step": 40458
    },
    {
      "epoch": 9.631315836457775,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9896,
      "step": 40459
    },
    {
      "epoch": 9.631553889186454,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 40460
    },
    {
      "epoch": 9.631791941915134,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8015,
      "step": 40461
    },
    {
      "epoch": 9.632029994643814,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.7686,
      "step": 40462
    },
    {
      "epoch": 9.632268047372493,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1351,
      "step": 40463
    },
    {
      "epoch": 9.632506100101173,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0551,
      "step": 40464
    },
    {
      "epoch": 9.632744152829853,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8126,
      "step": 40465
    },
    {
      "epoch": 9.63298220555853,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 2.1204,
      "step": 40466
    },
    {
      "epoch": 9.63322025828721,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 40467
    },
    {
      "epoch": 9.63345831101589,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.832,
      "step": 40468
    },
    {
      "epoch": 9.63369636374457,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.4791,
      "step": 40469
    },
    {
      "epoch": 9.633934416473249,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2624,
      "step": 40470
    },
    {
      "epoch": 9.634172469201928,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.8495,
      "step": 40471
    },
    {
      "epoch": 9.634410521930608,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0348,
      "step": 40472
    },
    {
      "epoch": 9.634648574659288,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.111,
      "step": 40473
    },
    {
      "epoch": 9.634886627387967,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.014,
      "step": 40474
    },
    {
      "epoch": 9.635124680116645,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 40475
    },
    {
      "epoch": 9.635362732845325,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9548,
      "step": 40476
    },
    {
      "epoch": 9.635600785574004,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 40477
    },
    {
      "epoch": 9.635838838302684,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.8295,
      "step": 40478
    },
    {
      "epoch": 9.636076891031363,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0626,
      "step": 40479
    },
    {
      "epoch": 9.636314943760043,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8302,
      "step": 40480
    },
    {
      "epoch": 9.636552996488723,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8652,
      "step": 40481
    },
    {
      "epoch": 9.636791049217402,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0158,
      "step": 40482
    },
    {
      "epoch": 9.637029101946082,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0721,
      "step": 40483
    },
    {
      "epoch": 9.63726715467476,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0207,
      "step": 40484
    },
    {
      "epoch": 9.63750520740344,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8802,
      "step": 40485
    },
    {
      "epoch": 9.637743260132119,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.979,
      "step": 40486
    },
    {
      "epoch": 9.637981312860799,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.2646,
      "step": 40487
    },
    {
      "epoch": 9.638219365589478,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1276,
      "step": 40488
    },
    {
      "epoch": 9.638457418318158,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0765,
      "step": 40489
    },
    {
      "epoch": 9.638695471046837,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2147,
      "step": 40490
    },
    {
      "epoch": 9.638933523775517,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.8504,
      "step": 40491
    },
    {
      "epoch": 9.639171576504197,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 40492
    },
    {
      "epoch": 9.639409629232874,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 40493
    },
    {
      "epoch": 9.639647681961554,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 40494
    },
    {
      "epoch": 9.639885734690234,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9307,
      "step": 40495
    },
    {
      "epoch": 9.640123787418913,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.8853,
      "step": 40496
    },
    {
      "epoch": 9.640361840147593,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9852,
      "step": 40497
    },
    {
      "epoch": 9.640599892876272,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0232,
      "step": 40498
    },
    {
      "epoch": 9.640837945604952,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8287,
      "step": 40499
    },
    {
      "epoch": 9.641075998333632,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9794,
      "step": 40500
    },
    {
      "epoch": 9.641075998333632,
      "eval_runtime": 0.524,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 1.908,
      "step": 40500
    },
    {
      "epoch": 9.64131405106231,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1366,
      "step": 40501
    },
    {
      "epoch": 9.641552103790989,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9761,
      "step": 40502
    },
    {
      "epoch": 9.641790156519669,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0751,
      "step": 40503
    },
    {
      "epoch": 9.642028209248348,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1085,
      "step": 40504
    },
    {
      "epoch": 9.642266261977028,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.7925,
      "step": 40505
    },
    {
      "epoch": 9.642504314705707,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9085,
      "step": 40506
    },
    {
      "epoch": 9.642742367434387,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 2.0843,
      "step": 40507
    },
    {
      "epoch": 9.642980420163067,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0496,
      "step": 40508
    },
    {
      "epoch": 9.643218472891746,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.1396,
      "step": 40509
    },
    {
      "epoch": 9.643456525620424,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0078,
      "step": 40510
    },
    {
      "epoch": 9.643694578349104,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1748,
      "step": 40511
    },
    {
      "epoch": 9.643932631077783,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 40512
    },
    {
      "epoch": 9.644170683806463,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9559,
      "step": 40513
    },
    {
      "epoch": 9.644408736535143,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.2322,
      "step": 40514
    },
    {
      "epoch": 9.644646789263822,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1841,
      "step": 40515
    },
    {
      "epoch": 9.644884841992502,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.7373,
      "step": 40516
    },
    {
      "epoch": 9.645122894721181,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 40517
    },
    {
      "epoch": 9.645360947449861,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1421,
      "step": 40518
    },
    {
      "epoch": 9.645599000178539,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.002,
      "step": 40519
    },
    {
      "epoch": 9.645837052907218,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3158,
      "step": 40520
    },
    {
      "epoch": 9.646075105635898,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1278,
      "step": 40521
    },
    {
      "epoch": 9.646313158364578,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.1626,
      "step": 40522
    },
    {
      "epoch": 9.646551211093257,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 40523
    },
    {
      "epoch": 9.646789263821937,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 40524
    },
    {
      "epoch": 9.647027316550616,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 40525
    },
    {
      "epoch": 9.647265369279296,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0424,
      "step": 40526
    },
    {
      "epoch": 9.647503422007976,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0988,
      "step": 40527
    },
    {
      "epoch": 9.647741474736653,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.052,
      "step": 40528
    },
    {
      "epoch": 9.647979527465333,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 40529
    },
    {
      "epoch": 9.648217580194013,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.937,
      "step": 40530
    },
    {
      "epoch": 9.648455632922692,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 40531
    },
    {
      "epoch": 9.648693685651372,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8648,
      "step": 40532
    },
    {
      "epoch": 9.648931738380051,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.2754,
      "step": 40533
    },
    {
      "epoch": 9.649169791108731,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.039,
      "step": 40534
    },
    {
      "epoch": 9.64940784383741,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.7563,
      "step": 40535
    },
    {
      "epoch": 9.649645896566089,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1688,
      "step": 40536
    },
    {
      "epoch": 9.649883949294768,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0825,
      "step": 40537
    },
    {
      "epoch": 9.650122002023448,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8464,
      "step": 40538
    },
    {
      "epoch": 9.650360054752127,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.1964,
      "step": 40539
    },
    {
      "epoch": 9.650598107480807,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8552,
      "step": 40540
    },
    {
      "epoch": 9.650836160209487,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0539,
      "step": 40541
    },
    {
      "epoch": 9.651074212938166,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9677,
      "step": 40542
    },
    {
      "epoch": 9.651312265666846,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 1.7375,
      "step": 40543
    },
    {
      "epoch": 9.651550318395525,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.3069,
      "step": 40544
    },
    {
      "epoch": 9.651788371124203,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0353,
      "step": 40545
    },
    {
      "epoch": 9.652026423852883,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2223,
      "step": 40546
    },
    {
      "epoch": 9.652264476581562,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9969,
      "step": 40547
    },
    {
      "epoch": 9.652502529310242,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2616,
      "step": 40548
    },
    {
      "epoch": 9.652740582038922,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.8286,
      "step": 40549
    },
    {
      "epoch": 9.652978634767601,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7276,
      "step": 40550
    },
    {
      "epoch": 9.65321668749628,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 40551
    },
    {
      "epoch": 9.65345474022496,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.2054,
      "step": 40552
    },
    {
      "epoch": 9.65369279295364,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 40553
    },
    {
      "epoch": 9.653930845682318,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.985,
      "step": 40554
    },
    {
      "epoch": 9.654168898410997,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0499,
      "step": 40555
    },
    {
      "epoch": 9.654406951139677,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0395,
      "step": 40556
    },
    {
      "epoch": 9.654645003868357,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 40557
    },
    {
      "epoch": 9.654883056597036,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1849,
      "step": 40558
    },
    {
      "epoch": 9.655121109325716,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 40559
    },
    {
      "epoch": 9.655359162054395,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9143,
      "step": 40560
    },
    {
      "epoch": 9.655597214783075,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9763,
      "step": 40561
    },
    {
      "epoch": 9.655835267511755,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9447,
      "step": 40562
    },
    {
      "epoch": 9.656073320240433,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.2573,
      "step": 40563
    },
    {
      "epoch": 9.656311372969112,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.817,
      "step": 40564
    },
    {
      "epoch": 9.656549425697792,
      "grad_norm": 5.5,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 40565
    },
    {
      "epoch": 9.656787478426471,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0707,
      "step": 40566
    },
    {
      "epoch": 9.657025531155151,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8994,
      "step": 40567
    },
    {
      "epoch": 9.65726358388383,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.215,
      "step": 40568
    },
    {
      "epoch": 9.65750163661251,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 40569
    },
    {
      "epoch": 9.65773968934119,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9796,
      "step": 40570
    },
    {
      "epoch": 9.657977742069868,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9607,
      "step": 40571
    },
    {
      "epoch": 9.658215794798547,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2034,
      "step": 40572
    },
    {
      "epoch": 9.658453847527227,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1361,
      "step": 40573
    },
    {
      "epoch": 9.658691900255906,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.0155,
      "step": 40574
    },
    {
      "epoch": 9.658929952984586,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.088,
      "step": 40575
    },
    {
      "epoch": 9.659168005713266,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 40576
    },
    {
      "epoch": 9.659406058441945,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1216,
      "step": 40577
    },
    {
      "epoch": 9.659644111170625,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 40578
    },
    {
      "epoch": 9.659882163899304,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.6522,
      "step": 40579
    },
    {
      "epoch": 9.660120216627984,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 40580
    },
    {
      "epoch": 9.660358269356662,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1483,
      "step": 40581
    },
    {
      "epoch": 9.660596322085341,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9125,
      "step": 40582
    },
    {
      "epoch": 9.660834374814021,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 40583
    },
    {
      "epoch": 9.6610724275427,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8352,
      "step": 40584
    },
    {
      "epoch": 9.66131048027138,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 40585
    },
    {
      "epoch": 9.66154853300006,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.933,
      "step": 40586
    },
    {
      "epoch": 9.66178658572874,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9842,
      "step": 40587
    },
    {
      "epoch": 9.662024638457419,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.9299,
      "step": 40588
    },
    {
      "epoch": 9.662262691186097,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 40589
    },
    {
      "epoch": 9.662500743914777,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8386,
      "step": 40590
    },
    {
      "epoch": 9.662738796643456,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9121,
      "step": 40591
    },
    {
      "epoch": 9.662976849372136,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0346,
      "step": 40592
    },
    {
      "epoch": 9.663214902100815,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.187,
      "step": 40593
    },
    {
      "epoch": 9.663452954829495,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3218,
      "step": 40594
    },
    {
      "epoch": 9.663691007558175,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8431,
      "step": 40595
    },
    {
      "epoch": 9.663929060286854,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9765,
      "step": 40596
    },
    {
      "epoch": 9.664167113015534,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 40597
    },
    {
      "epoch": 9.664405165744212,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0115,
      "step": 40598
    },
    {
      "epoch": 9.664643218472891,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9985,
      "step": 40599
    },
    {
      "epoch": 9.66488127120157,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8525,
      "step": 40600
    },
    {
      "epoch": 9.66488127120157,
      "eval_runtime": 0.5299,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 40600
    },
    {
      "epoch": 9.66511932393025,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.804,
      "step": 40601
    },
    {
      "epoch": 9.66535737665893,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9623,
      "step": 40602
    },
    {
      "epoch": 9.66559542938761,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.5018,
      "step": 40603
    },
    {
      "epoch": 9.66583348211629,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 40604
    },
    {
      "epoch": 9.666071534844969,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1454,
      "step": 40605
    },
    {
      "epoch": 9.666309587573647,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2429,
      "step": 40606
    },
    {
      "epoch": 9.666547640302326,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9459,
      "step": 40607
    },
    {
      "epoch": 9.666785693031006,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2491,
      "step": 40608
    },
    {
      "epoch": 9.667023745759685,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 40609
    },
    {
      "epoch": 9.667261798488365,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9683,
      "step": 40610
    },
    {
      "epoch": 9.667499851217045,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.8327,
      "step": 40611
    },
    {
      "epoch": 9.667737903945724,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8632,
      "step": 40612
    },
    {
      "epoch": 9.667975956674404,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 2.1945,
      "step": 40613
    },
    {
      "epoch": 9.668214009403084,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0877,
      "step": 40614
    },
    {
      "epoch": 9.668452062131763,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1447,
      "step": 40615
    },
    {
      "epoch": 9.668690114860441,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1837,
      "step": 40616
    },
    {
      "epoch": 9.66892816758912,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3214,
      "step": 40617
    },
    {
      "epoch": 9.6691662203178,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.0558,
      "step": 40618
    },
    {
      "epoch": 9.66940427304648,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0342,
      "step": 40619
    },
    {
      "epoch": 9.66964232577516,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2007,
      "step": 40620
    },
    {
      "epoch": 9.669880378503839,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.7059,
      "step": 40621
    },
    {
      "epoch": 9.670118431232519,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1217,
      "step": 40622
    },
    {
      "epoch": 9.670356483961198,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0743,
      "step": 40623
    },
    {
      "epoch": 9.670594536689876,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2623,
      "step": 40624
    },
    {
      "epoch": 9.670832589418556,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0111,
      "step": 40625
    },
    {
      "epoch": 9.671070642147235,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0909,
      "step": 40626
    },
    {
      "epoch": 9.671308694875915,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9477,
      "step": 40627
    },
    {
      "epoch": 9.671546747604594,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9008,
      "step": 40628
    },
    {
      "epoch": 9.671784800333274,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.0199,
      "step": 40629
    },
    {
      "epoch": 9.672022853061954,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0512,
      "step": 40630
    },
    {
      "epoch": 9.672260905790633,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8884,
      "step": 40631
    },
    {
      "epoch": 9.672498958519313,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.974,
      "step": 40632
    },
    {
      "epoch": 9.67273701124799,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0652,
      "step": 40633
    },
    {
      "epoch": 9.67297506397667,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.3823,
      "step": 40634
    },
    {
      "epoch": 9.67321311670535,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 40635
    },
    {
      "epoch": 9.67345116943403,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.8899,
      "step": 40636
    },
    {
      "epoch": 9.673689222162709,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.939,
      "step": 40637
    },
    {
      "epoch": 9.673927274891389,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9648,
      "step": 40638
    },
    {
      "epoch": 9.674165327620068,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9394,
      "step": 40639
    },
    {
      "epoch": 9.674403380348748,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 40640
    },
    {
      "epoch": 9.674641433077426,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7313,
      "step": 40641
    },
    {
      "epoch": 9.674879485806105,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2226,
      "step": 40642
    },
    {
      "epoch": 9.675117538534785,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2589,
      "step": 40643
    },
    {
      "epoch": 9.675355591263465,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9179,
      "step": 40644
    },
    {
      "epoch": 9.675593643992144,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2484,
      "step": 40645
    },
    {
      "epoch": 9.675831696720824,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.347,
      "step": 40646
    },
    {
      "epoch": 9.676069749449503,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.8569,
      "step": 40647
    },
    {
      "epoch": 9.676307802178183,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 2.1424,
      "step": 40648
    },
    {
      "epoch": 9.676545854906863,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1083,
      "step": 40649
    },
    {
      "epoch": 9.676783907635542,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9389,
      "step": 40650
    },
    {
      "epoch": 9.67702196036422,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8475,
      "step": 40651
    },
    {
      "epoch": 9.6772600130929,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.7884,
      "step": 40652
    },
    {
      "epoch": 9.67749806582158,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 40653
    },
    {
      "epoch": 9.677736118550259,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.0557,
      "step": 40654
    },
    {
      "epoch": 9.677974171278938,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 40655
    },
    {
      "epoch": 9.678212224007618,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0298,
      "step": 40656
    },
    {
      "epoch": 9.678450276736298,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 40657
    },
    {
      "epoch": 9.678688329464977,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.062,
      "step": 40658
    },
    {
      "epoch": 9.678926382193655,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0748,
      "step": 40659
    },
    {
      "epoch": 9.679164434922335,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.9345,
      "step": 40660
    },
    {
      "epoch": 9.679402487651014,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1718,
      "step": 40661
    },
    {
      "epoch": 9.679640540379694,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.258,
      "step": 40662
    },
    {
      "epoch": 9.679878593108374,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.8597,
      "step": 40663
    },
    {
      "epoch": 9.680116645837053,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1109,
      "step": 40664
    },
    {
      "epoch": 9.680354698565733,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9296,
      "step": 40665
    },
    {
      "epoch": 9.680592751294412,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.9578,
      "step": 40666
    },
    {
      "epoch": 9.680830804023092,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 1.9048,
      "step": 40667
    },
    {
      "epoch": 9.68106885675177,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.8463,
      "step": 40668
    },
    {
      "epoch": 9.68130690948045,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 40669
    },
    {
      "epoch": 9.681544962209129,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.1176,
      "step": 40670
    },
    {
      "epoch": 9.681783014937809,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9416,
      "step": 40671
    },
    {
      "epoch": 9.682021067666488,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 40672
    },
    {
      "epoch": 9.682259120395168,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.2479,
      "step": 40673
    },
    {
      "epoch": 9.682497173123847,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 1.8797,
      "step": 40674
    },
    {
      "epoch": 9.682735225852527,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 1.9237,
      "step": 40675
    },
    {
      "epoch": 9.682973278581205,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 40676
    },
    {
      "epoch": 9.683211331309884,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.7796,
      "step": 40677
    },
    {
      "epoch": 9.683449384038564,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0237,
      "step": 40678
    },
    {
      "epoch": 9.683687436767244,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9422,
      "step": 40679
    },
    {
      "epoch": 9.683925489495923,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8784,
      "step": 40680
    },
    {
      "epoch": 9.684163542224603,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8119,
      "step": 40681
    },
    {
      "epoch": 9.684401594953282,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9231,
      "step": 40682
    },
    {
      "epoch": 9.684639647681962,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9265,
      "step": 40683
    },
    {
      "epoch": 9.684877700410642,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0281,
      "step": 40684
    },
    {
      "epoch": 9.685115753139321,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.8787,
      "step": 40685
    },
    {
      "epoch": 9.685353805867999,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1215,
      "step": 40686
    },
    {
      "epoch": 9.685591858596679,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.0544,
      "step": 40687
    },
    {
      "epoch": 9.685829911325358,
      "grad_norm": 3.0625,
      "learning_rate": 3e-05,
      "loss": 1.7322,
      "step": 40688
    },
    {
      "epoch": 9.686067964054038,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0242,
      "step": 40689
    },
    {
      "epoch": 9.686306016782718,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9601,
      "step": 40690
    },
    {
      "epoch": 9.686544069511397,
      "grad_norm": 2.921875,
      "learning_rate": 3e-05,
      "loss": 1.984,
      "step": 40691
    },
    {
      "epoch": 9.686782122240077,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0285,
      "step": 40692
    },
    {
      "epoch": 9.687020174968756,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.0547,
      "step": 40693
    },
    {
      "epoch": 9.687258227697434,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.1427,
      "step": 40694
    },
    {
      "epoch": 9.687496280426114,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.003,
      "step": 40695
    },
    {
      "epoch": 9.687734333154793,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.2603,
      "step": 40696
    },
    {
      "epoch": 9.687972385883473,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0865,
      "step": 40697
    },
    {
      "epoch": 9.688210438612153,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9169,
      "step": 40698
    },
    {
      "epoch": 9.688448491340832,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 2.0131,
      "step": 40699
    },
    {
      "epoch": 9.688686544069512,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 40700
    },
    {
      "epoch": 9.688686544069512,
      "eval_runtime": 0.5299,
      "eval_samples_per_second": 1.887,
      "eval_steps_per_second": 1.887,
      "step": 40700
    },
    {
      "epoch": 9.688924596798191,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9813,
      "step": 40701
    },
    {
      "epoch": 9.689162649526871,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.7685,
      "step": 40702
    },
    {
      "epoch": 9.689400702255549,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 40703
    },
    {
      "epoch": 9.689638754984228,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0385,
      "step": 40704
    },
    {
      "epoch": 9.689876807712908,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.149,
      "step": 40705
    },
    {
      "epoch": 9.690114860441588,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9726,
      "step": 40706
    },
    {
      "epoch": 9.690352913170267,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.7768,
      "step": 40707
    },
    {
      "epoch": 9.690590965898947,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.7726,
      "step": 40708
    },
    {
      "epoch": 9.690829018627626,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2319,
      "step": 40709
    },
    {
      "epoch": 9.691067071356306,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 40710
    },
    {
      "epoch": 9.691305124084984,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.1129,
      "step": 40711
    },
    {
      "epoch": 9.691543176813664,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9471,
      "step": 40712
    },
    {
      "epoch": 9.691781229542343,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 40713
    },
    {
      "epoch": 9.692019282271023,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0286,
      "step": 40714
    },
    {
      "epoch": 9.692257334999702,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8685,
      "step": 40715
    },
    {
      "epoch": 9.692495387728382,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.2209,
      "step": 40716
    },
    {
      "epoch": 9.692733440457062,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.9306,
      "step": 40717
    },
    {
      "epoch": 9.692971493185741,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9502,
      "step": 40718
    },
    {
      "epoch": 9.69320954591442,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0056,
      "step": 40719
    },
    {
      "epoch": 9.6934475986431,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 40720
    },
    {
      "epoch": 9.693685651371778,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1092,
      "step": 40721
    },
    {
      "epoch": 9.693923704100458,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9052,
      "step": 40722
    },
    {
      "epoch": 9.694161756829137,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9558,
      "step": 40723
    },
    {
      "epoch": 9.694399809557817,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8412,
      "step": 40724
    },
    {
      "epoch": 9.694637862286497,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.1049,
      "step": 40725
    },
    {
      "epoch": 9.694875915015176,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0432,
      "step": 40726
    },
    {
      "epoch": 9.695113967743856,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0357,
      "step": 40727
    },
    {
      "epoch": 9.695352020472535,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.1788,
      "step": 40728
    },
    {
      "epoch": 9.695590073201213,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3522,
      "step": 40729
    },
    {
      "epoch": 9.695828125929893,
      "grad_norm": 3.25,
      "learning_rate": 3e-05,
      "loss": 1.7964,
      "step": 40730
    },
    {
      "epoch": 9.696066178658572,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.1666,
      "step": 40731
    },
    {
      "epoch": 9.696304231387252,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9517,
      "step": 40732
    },
    {
      "epoch": 9.696542284115932,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 40733
    },
    {
      "epoch": 9.696780336844611,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.825,
      "step": 40734
    },
    {
      "epoch": 9.69701838957329,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.013,
      "step": 40735
    },
    {
      "epoch": 9.69725644230197,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.3434,
      "step": 40736
    },
    {
      "epoch": 9.69749449503065,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.8796,
      "step": 40737
    },
    {
      "epoch": 9.697732547759328,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 40738
    },
    {
      "epoch": 9.697970600488008,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0844,
      "step": 40739
    },
    {
      "epoch": 9.698208653216687,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9273,
      "step": 40740
    },
    {
      "epoch": 9.698446705945367,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 40741
    },
    {
      "epoch": 9.698684758674046,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8261,
      "step": 40742
    },
    {
      "epoch": 9.698922811402726,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9429,
      "step": 40743
    },
    {
      "epoch": 9.699160864131406,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.829,
      "step": 40744
    },
    {
      "epoch": 9.699398916860085,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9535,
      "step": 40745
    },
    {
      "epoch": 9.699636969588763,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.2124,
      "step": 40746
    },
    {
      "epoch": 9.699875022317443,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7259,
      "step": 40747
    },
    {
      "epoch": 9.700113075046122,
      "grad_norm": 2.890625,
      "learning_rate": 3e-05,
      "loss": 2.1642,
      "step": 40748
    },
    {
      "epoch": 9.700351127774802,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9585,
      "step": 40749
    },
    {
      "epoch": 9.700589180503481,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0383,
      "step": 40750
    },
    {
      "epoch": 9.700827233232161,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0887,
      "step": 40751
    },
    {
      "epoch": 9.70106528596084,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2117,
      "step": 40752
    },
    {
      "epoch": 9.70130333868952,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2154,
      "step": 40753
    },
    {
      "epoch": 9.7015413914182,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0888,
      "step": 40754
    },
    {
      "epoch": 9.70177944414688,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8877,
      "step": 40755
    },
    {
      "epoch": 9.702017496875557,
      "grad_norm": 3.03125,
      "learning_rate": 3e-05,
      "loss": 1.9597,
      "step": 40756
    },
    {
      "epoch": 9.702255549604237,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 40757
    },
    {
      "epoch": 9.702493602332916,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2366,
      "step": 40758
    },
    {
      "epoch": 9.702731655061596,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.2396,
      "step": 40759
    },
    {
      "epoch": 9.702969707790276,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9361,
      "step": 40760
    },
    {
      "epoch": 9.703207760518955,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 40761
    },
    {
      "epoch": 9.703445813247635,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.1654,
      "step": 40762
    },
    {
      "epoch": 9.703683865976314,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0447,
      "step": 40763
    },
    {
      "epoch": 9.703921918704992,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.3175,
      "step": 40764
    },
    {
      "epoch": 9.704159971433672,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1374,
      "step": 40765
    },
    {
      "epoch": 9.704398024162352,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9822,
      "step": 40766
    },
    {
      "epoch": 9.704636076891031,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0423,
      "step": 40767
    },
    {
      "epoch": 9.70487412961971,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1467,
      "step": 40768
    },
    {
      "epoch": 9.70511218234839,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 40769
    },
    {
      "epoch": 9.70535023507707,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.9421,
      "step": 40770
    },
    {
      "epoch": 9.70558828780575,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.102,
      "step": 40771
    },
    {
      "epoch": 9.70582634053443,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.7899,
      "step": 40772
    },
    {
      "epoch": 9.706064393263107,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8716,
      "step": 40773
    },
    {
      "epoch": 9.706302445991787,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.9553,
      "step": 40774
    },
    {
      "epoch": 9.706540498720466,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.3415,
      "step": 40775
    },
    {
      "epoch": 9.706778551449146,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2609,
      "step": 40776
    },
    {
      "epoch": 9.707016604177825,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.3307,
      "step": 40777
    },
    {
      "epoch": 9.707254656906505,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.8377,
      "step": 40778
    },
    {
      "epoch": 9.707492709635185,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1325,
      "step": 40779
    },
    {
      "epoch": 9.707730762363864,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9316,
      "step": 40780
    },
    {
      "epoch": 9.707968815092542,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.7151,
      "step": 40781
    },
    {
      "epoch": 9.708206867821222,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.0886,
      "step": 40782
    },
    {
      "epoch": 9.708444920549901,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.2165,
      "step": 40783
    },
    {
      "epoch": 9.70868297327858,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1858,
      "step": 40784
    },
    {
      "epoch": 9.70892102600726,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0814,
      "step": 40785
    },
    {
      "epoch": 9.70915907873594,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.7986,
      "step": 40786
    },
    {
      "epoch": 9.70939713146462,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.146,
      "step": 40787
    },
    {
      "epoch": 9.7096351841933,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1701,
      "step": 40788
    },
    {
      "epoch": 9.709873236921979,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0646,
      "step": 40789
    },
    {
      "epoch": 9.710111289650659,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 2.2583,
      "step": 40790
    },
    {
      "epoch": 9.710349342379336,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.9521,
      "step": 40791
    },
    {
      "epoch": 9.710587395108016,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1961,
      "step": 40792
    },
    {
      "epoch": 9.710825447836696,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 2.0642,
      "step": 40793
    },
    {
      "epoch": 9.711063500565375,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9226,
      "step": 40794
    },
    {
      "epoch": 9.711301553294055,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.075,
      "step": 40795
    },
    {
      "epoch": 9.711539606022734,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.848,
      "step": 40796
    },
    {
      "epoch": 9.711777658751414,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8174,
      "step": 40797
    },
    {
      "epoch": 9.712015711480094,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0469,
      "step": 40798
    },
    {
      "epoch": 9.712253764208771,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.1491,
      "step": 40799
    },
    {
      "epoch": 9.712491816937451,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.1321,
      "step": 40800
    },
    {
      "epoch": 9.712491816937451,
      "eval_runtime": 0.5288,
      "eval_samples_per_second": 1.891,
      "eval_steps_per_second": 1.891,
      "step": 40800
    },
    {
      "epoch": 9.71272986966613,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.1135,
      "step": 40801
    },
    {
      "epoch": 9.71296792239481,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.168,
      "step": 40802
    },
    {
      "epoch": 9.71320597512349,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 40803
    },
    {
      "epoch": 9.71344402785217,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9251,
      "step": 40804
    },
    {
      "epoch": 9.713682080580849,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8593,
      "step": 40805
    },
    {
      "epoch": 9.713920133309529,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0205,
      "step": 40806
    },
    {
      "epoch": 9.714158186038208,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7225,
      "step": 40807
    },
    {
      "epoch": 9.714396238766886,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.07,
      "step": 40808
    },
    {
      "epoch": 9.714634291495566,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9758,
      "step": 40809
    },
    {
      "epoch": 9.714872344224245,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0484,
      "step": 40810
    },
    {
      "epoch": 9.715110396952925,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.2935,
      "step": 40811
    },
    {
      "epoch": 9.715348449681604,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0747,
      "step": 40812
    },
    {
      "epoch": 9.715586502410284,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9469,
      "step": 40813
    },
    {
      "epoch": 9.715824555138964,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 40814
    },
    {
      "epoch": 9.716062607867643,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0197,
      "step": 40815
    },
    {
      "epoch": 9.716300660596323,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9339,
      "step": 40816
    },
    {
      "epoch": 9.716538713325,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3341,
      "step": 40817
    },
    {
      "epoch": 9.71677676605368,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.0426,
      "step": 40818
    },
    {
      "epoch": 9.71701481878236,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1324,
      "step": 40819
    },
    {
      "epoch": 9.71725287151104,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0308,
      "step": 40820
    },
    {
      "epoch": 9.71749092423972,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 40821
    },
    {
      "epoch": 9.717728976968399,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8768,
      "step": 40822
    },
    {
      "epoch": 9.717967029697078,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.107,
      "step": 40823
    },
    {
      "epoch": 9.718205082425758,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.7911,
      "step": 40824
    },
    {
      "epoch": 9.718443135154438,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 2.2277,
      "step": 40825
    },
    {
      "epoch": 9.718681187883115,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.2665,
      "step": 40826
    },
    {
      "epoch": 9.718919240611795,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 40827
    },
    {
      "epoch": 9.719157293340475,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8388,
      "step": 40828
    },
    {
      "epoch": 9.719395346069154,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.161,
      "step": 40829
    },
    {
      "epoch": 9.719633398797834,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9909,
      "step": 40830
    },
    {
      "epoch": 9.719871451526513,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 40831
    },
    {
      "epoch": 9.720109504255193,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9537,
      "step": 40832
    },
    {
      "epoch": 9.720347556983873,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.9082,
      "step": 40833
    },
    {
      "epoch": 9.72058560971255,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8799,
      "step": 40834
    },
    {
      "epoch": 9.72082366244123,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.965,
      "step": 40835
    },
    {
      "epoch": 9.72106171516991,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.2379,
      "step": 40836
    },
    {
      "epoch": 9.72129976789859,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 40837
    },
    {
      "epoch": 9.721537820627269,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 40838
    },
    {
      "epoch": 9.721775873355949,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1533,
      "step": 40839
    },
    {
      "epoch": 9.722013926084628,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.036,
      "step": 40840
    },
    {
      "epoch": 9.722251978813308,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 40841
    },
    {
      "epoch": 9.722490031541987,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8841,
      "step": 40842
    },
    {
      "epoch": 9.722728084270665,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.5263,
      "step": 40843
    },
    {
      "epoch": 9.722966136999345,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.2845,
      "step": 40844
    },
    {
      "epoch": 9.723204189728024,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1894,
      "step": 40845
    },
    {
      "epoch": 9.723442242456704,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 40846
    },
    {
      "epoch": 9.723680295185384,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8949,
      "step": 40847
    },
    {
      "epoch": 9.723918347914063,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.93,
      "step": 40848
    },
    {
      "epoch": 9.724156400642743,
      "grad_norm": 6.03125,
      "learning_rate": 3e-05,
      "loss": 1.7626,
      "step": 40849
    },
    {
      "epoch": 9.724394453371422,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.0435,
      "step": 40850
    },
    {
      "epoch": 9.724632506100102,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8604,
      "step": 40851
    },
    {
      "epoch": 9.72487055882878,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9891,
      "step": 40852
    },
    {
      "epoch": 9.72510861155746,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9168,
      "step": 40853
    },
    {
      "epoch": 9.725346664286139,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.948,
      "step": 40854
    },
    {
      "epoch": 9.725584717014819,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.1156,
      "step": 40855
    },
    {
      "epoch": 9.725822769743498,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1684,
      "step": 40856
    },
    {
      "epoch": 9.726060822472178,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.7915,
      "step": 40857
    },
    {
      "epoch": 9.726298875200857,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0421,
      "step": 40858
    },
    {
      "epoch": 9.726536927929537,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1969,
      "step": 40859
    },
    {
      "epoch": 9.726774980658217,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 40860
    },
    {
      "epoch": 9.727013033386894,
      "grad_norm": 4.65625,
      "learning_rate": 3e-05,
      "loss": 1.805,
      "step": 40861
    },
    {
      "epoch": 9.727251086115574,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0289,
      "step": 40862
    },
    {
      "epoch": 9.727489138844254,
      "grad_norm": 5.25,
      "learning_rate": 3e-05,
      "loss": 2.0591,
      "step": 40863
    },
    {
      "epoch": 9.727727191572933,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.841,
      "step": 40864
    },
    {
      "epoch": 9.727965244301613,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9995,
      "step": 40865
    },
    {
      "epoch": 9.728203297030293,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.7405,
      "step": 40866
    },
    {
      "epoch": 9.728441349758972,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.872,
      "step": 40867
    },
    {
      "epoch": 9.728679402487652,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 40868
    },
    {
      "epoch": 9.72891745521633,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8622,
      "step": 40869
    },
    {
      "epoch": 9.72915550794501,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.9747,
      "step": 40870
    },
    {
      "epoch": 9.729393560673689,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1641,
      "step": 40871
    },
    {
      "epoch": 9.729631613402368,
      "grad_norm": 2.828125,
      "learning_rate": 3e-05,
      "loss": 1.8814,
      "step": 40872
    },
    {
      "epoch": 9.729869666131048,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.2373,
      "step": 40873
    },
    {
      "epoch": 9.730107718859728,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8944,
      "step": 40874
    },
    {
      "epoch": 9.730345771588407,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.4942,
      "step": 40875
    },
    {
      "epoch": 9.730583824317087,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1747,
      "step": 40876
    },
    {
      "epoch": 9.730821877045766,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1405,
      "step": 40877
    },
    {
      "epoch": 9.731059929774444,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0578,
      "step": 40878
    },
    {
      "epoch": 9.731297982503124,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 40879
    },
    {
      "epoch": 9.731536035231803,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9841,
      "step": 40880
    },
    {
      "epoch": 9.731774087960483,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 1.8507,
      "step": 40881
    },
    {
      "epoch": 9.732012140689163,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9778,
      "step": 40882
    },
    {
      "epoch": 9.732250193417842,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0422,
      "step": 40883
    },
    {
      "epoch": 9.732488246146522,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 40884
    },
    {
      "epoch": 9.732726298875201,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 1.9986,
      "step": 40885
    },
    {
      "epoch": 9.732964351603881,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7207,
      "step": 40886
    },
    {
      "epoch": 9.733202404332559,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.9433,
      "step": 40887
    },
    {
      "epoch": 9.733440457061239,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9218,
      "step": 40888
    },
    {
      "epoch": 9.733678509789918,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.2087,
      "step": 40889
    },
    {
      "epoch": 9.733916562518598,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.1617,
      "step": 40890
    },
    {
      "epoch": 9.734154615247277,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.0657,
      "step": 40891
    },
    {
      "epoch": 9.734392667975957,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0754,
      "step": 40892
    },
    {
      "epoch": 9.734630720704637,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.1477,
      "step": 40893
    },
    {
      "epoch": 9.734868773433316,
      "grad_norm": 5.0,
      "learning_rate": 3e-05,
      "loss": 2.3857,
      "step": 40894
    },
    {
      "epoch": 9.735106826161996,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.4564,
      "step": 40895
    },
    {
      "epoch": 9.735344878890674,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7466,
      "step": 40896
    },
    {
      "epoch": 9.735582931619353,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.2267,
      "step": 40897
    },
    {
      "epoch": 9.735820984348033,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9061,
      "step": 40898
    },
    {
      "epoch": 9.736059037076712,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9799,
      "step": 40899
    },
    {
      "epoch": 9.736297089805392,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0297,
      "step": 40900
    },
    {
      "epoch": 9.736297089805392,
      "eval_runtime": 0.5427,
      "eval_samples_per_second": 1.843,
      "eval_steps_per_second": 1.843,
      "step": 40900
    },
    {
      "epoch": 9.736535142534072,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9709,
      "step": 40901
    },
    {
      "epoch": 9.736773195262751,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.4702,
      "step": 40902
    },
    {
      "epoch": 9.73701124799143,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.2656,
      "step": 40903
    },
    {
      "epoch": 9.737249300720109,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2315,
      "step": 40904
    },
    {
      "epoch": 9.737487353448788,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.8193,
      "step": 40905
    },
    {
      "epoch": 9.737725406177468,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.9131,
      "step": 40906
    },
    {
      "epoch": 9.737963458906147,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9282,
      "step": 40907
    },
    {
      "epoch": 9.738201511634827,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 40908
    },
    {
      "epoch": 9.738439564363507,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.8756,
      "step": 40909
    },
    {
      "epoch": 9.738677617092186,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0347,
      "step": 40910
    },
    {
      "epoch": 9.738915669820866,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0398,
      "step": 40911
    },
    {
      "epoch": 9.739153722549545,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8459,
      "step": 40912
    },
    {
      "epoch": 9.739391775278223,
      "grad_norm": 5.28125,
      "learning_rate": 3e-05,
      "loss": 1.9428,
      "step": 40913
    },
    {
      "epoch": 9.739629828006903,
      "grad_norm": 2.953125,
      "learning_rate": 3e-05,
      "loss": 2.0963,
      "step": 40914
    },
    {
      "epoch": 9.739867880735583,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.083,
      "step": 40915
    },
    {
      "epoch": 9.740105933464262,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.1764,
      "step": 40916
    },
    {
      "epoch": 9.740343986192942,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.883,
      "step": 40917
    },
    {
      "epoch": 9.740582038921621,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9894,
      "step": 40918
    },
    {
      "epoch": 9.740820091650301,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 40919
    },
    {
      "epoch": 9.74105814437898,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.0701,
      "step": 40920
    },
    {
      "epoch": 9.74129619710766,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9737,
      "step": 40921
    },
    {
      "epoch": 9.741534249836338,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0103,
      "step": 40922
    },
    {
      "epoch": 9.741772302565018,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 40923
    },
    {
      "epoch": 9.742010355293697,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.8573,
      "step": 40924
    },
    {
      "epoch": 9.742248408022377,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0734,
      "step": 40925
    },
    {
      "epoch": 9.742486460751056,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.8968,
      "step": 40926
    },
    {
      "epoch": 9.742724513479736,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0691,
      "step": 40927
    },
    {
      "epoch": 9.742962566208416,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 1.9208,
      "step": 40928
    },
    {
      "epoch": 9.743200618937095,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1919,
      "step": 40929
    },
    {
      "epoch": 9.743438671665775,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.0803,
      "step": 40930
    },
    {
      "epoch": 9.743676724394453,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9324,
      "step": 40931
    },
    {
      "epoch": 9.743914777123132,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8113,
      "step": 40932
    },
    {
      "epoch": 9.744152829851812,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 40933
    },
    {
      "epoch": 9.744390882580491,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 40934
    },
    {
      "epoch": 9.744628935309171,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1084,
      "step": 40935
    },
    {
      "epoch": 9.74486698803785,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9188,
      "step": 40936
    },
    {
      "epoch": 9.74510504076653,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.2163,
      "step": 40937
    },
    {
      "epoch": 9.74534309349521,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7015,
      "step": 40938
    },
    {
      "epoch": 9.745581146223888,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0836,
      "step": 40939
    },
    {
      "epoch": 9.745819198952567,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8548,
      "step": 40940
    },
    {
      "epoch": 9.746057251681247,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7689,
      "step": 40941
    },
    {
      "epoch": 9.746295304409927,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 1.5945,
      "step": 40942
    },
    {
      "epoch": 9.746533357138606,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0062,
      "step": 40943
    },
    {
      "epoch": 9.746771409867286,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9612,
      "step": 40944
    },
    {
      "epoch": 9.747009462595965,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.1916,
      "step": 40945
    },
    {
      "epoch": 9.747247515324645,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1072,
      "step": 40946
    },
    {
      "epoch": 9.747485568053325,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 2.1546,
      "step": 40947
    },
    {
      "epoch": 9.747723620782002,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0165,
      "step": 40948
    },
    {
      "epoch": 9.747961673510682,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0901,
      "step": 40949
    },
    {
      "epoch": 9.748199726239362,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1397,
      "step": 40950
    },
    {
      "epoch": 9.748437778968041,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1696,
      "step": 40951
    },
    {
      "epoch": 9.74867583169672,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 40952
    },
    {
      "epoch": 9.7489138844254,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0744,
      "step": 40953
    },
    {
      "epoch": 9.74915193715408,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.9112,
      "step": 40954
    },
    {
      "epoch": 9.74938998988276,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.8981,
      "step": 40955
    },
    {
      "epoch": 9.74962804261144,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0135,
      "step": 40956
    },
    {
      "epoch": 9.749866095340117,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 40957
    },
    {
      "epoch": 9.750104148068797,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1656,
      "step": 40958
    },
    {
      "epoch": 9.750342200797476,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.7808,
      "step": 40959
    },
    {
      "epoch": 9.750580253526156,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0301,
      "step": 40960
    },
    {
      "epoch": 9.750818306254835,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0713,
      "step": 40961
    },
    {
      "epoch": 9.751056358983515,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0117,
      "step": 40962
    },
    {
      "epoch": 9.751294411712195,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 40963
    },
    {
      "epoch": 9.751532464440874,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0797,
      "step": 40964
    },
    {
      "epoch": 9.751770517169554,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0316,
      "step": 40965
    },
    {
      "epoch": 9.752008569898232,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.1305,
      "step": 40966
    },
    {
      "epoch": 9.752246622626911,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.9392,
      "step": 40967
    },
    {
      "epoch": 9.752484675355591,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.7668,
      "step": 40968
    },
    {
      "epoch": 9.75272272808427,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.8901,
      "step": 40969
    },
    {
      "epoch": 9.75296078081295,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0274,
      "step": 40970
    },
    {
      "epoch": 9.75319883354163,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.6429,
      "step": 40971
    },
    {
      "epoch": 9.75343688627031,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.2022,
      "step": 40972
    },
    {
      "epoch": 9.753674938998989,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9735,
      "step": 40973
    },
    {
      "epoch": 9.753912991727667,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.8855,
      "step": 40974
    },
    {
      "epoch": 9.754151044456346,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9272,
      "step": 40975
    },
    {
      "epoch": 9.754389097185026,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1126,
      "step": 40976
    },
    {
      "epoch": 9.754627149913706,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.1041,
      "step": 40977
    },
    {
      "epoch": 9.754865202642385,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8512,
      "step": 40978
    },
    {
      "epoch": 9.755103255371065,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9224,
      "step": 40979
    },
    {
      "epoch": 9.755341308099744,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1699,
      "step": 40980
    },
    {
      "epoch": 9.755579360828424,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0095,
      "step": 40981
    },
    {
      "epoch": 9.755817413557104,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.1542,
      "step": 40982
    },
    {
      "epoch": 9.756055466285783,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.023,
      "step": 40983
    },
    {
      "epoch": 9.756293519014461,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0157,
      "step": 40984
    },
    {
      "epoch": 9.75653157174314,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.3176,
      "step": 40985
    },
    {
      "epoch": 9.75676962447182,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.0712,
      "step": 40986
    },
    {
      "epoch": 9.7570076772005,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 1.9146,
      "step": 40987
    },
    {
      "epoch": 9.75724572992918,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.9659,
      "step": 40988
    },
    {
      "epoch": 9.757483782657859,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8052,
      "step": 40989
    },
    {
      "epoch": 9.757721835386539,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 40990
    },
    {
      "epoch": 9.757959888115218,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0014,
      "step": 40991
    },
    {
      "epoch": 9.758197940843896,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.9902,
      "step": 40992
    },
    {
      "epoch": 9.758435993572576,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.5357,
      "step": 40993
    },
    {
      "epoch": 9.758674046301255,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.3159,
      "step": 40994
    },
    {
      "epoch": 9.758912099029935,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.1082,
      "step": 40995
    },
    {
      "epoch": 9.759150151758615,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 40996
    },
    {
      "epoch": 9.759388204487294,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.0987,
      "step": 40997
    },
    {
      "epoch": 9.759626257215974,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0611,
      "step": 40998
    },
    {
      "epoch": 9.759864309944653,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9779,
      "step": 40999
    },
    {
      "epoch": 9.760102362673333,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2978,
      "step": 41000
    },
    {
      "epoch": 9.760102362673333,
      "eval_runtime": 0.5339,
      "eval_samples_per_second": 1.873,
      "eval_steps_per_second": 1.873,
      "step": 41000
    },
    {
      "epoch": 9.76034041540201,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9262,
      "step": 41001
    },
    {
      "epoch": 9.76057846813069,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9849,
      "step": 41002
    },
    {
      "epoch": 9.76081652085937,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0834,
      "step": 41003
    },
    {
      "epoch": 9.76105457358805,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.3195,
      "step": 41004
    },
    {
      "epoch": 9.76129262631673,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9395,
      "step": 41005
    },
    {
      "epoch": 9.761530679045409,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.0064,
      "step": 41006
    },
    {
      "epoch": 9.761768731774088,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.117,
      "step": 41007
    },
    {
      "epoch": 9.762006784502768,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.0931,
      "step": 41008
    },
    {
      "epoch": 9.762244837231446,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0259,
      "step": 41009
    },
    {
      "epoch": 9.762482889960125,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.8509,
      "step": 41010
    },
    {
      "epoch": 9.762720942688805,
      "grad_norm": 3.234375,
      "learning_rate": 3e-05,
      "loss": 2.0399,
      "step": 41011
    },
    {
      "epoch": 9.762958995417485,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.077,
      "step": 41012
    },
    {
      "epoch": 9.763197048146164,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9495,
      "step": 41013
    },
    {
      "epoch": 9.763435100874844,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1373,
      "step": 41014
    },
    {
      "epoch": 9.763673153603524,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0546,
      "step": 41015
    },
    {
      "epoch": 9.763911206332203,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9105,
      "step": 41016
    },
    {
      "epoch": 9.764149259060883,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 41017
    },
    {
      "epoch": 9.764387311789562,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.2066,
      "step": 41018
    },
    {
      "epoch": 9.76462536451824,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.854,
      "step": 41019
    },
    {
      "epoch": 9.76486341724692,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8087,
      "step": 41020
    },
    {
      "epoch": 9.7651014699756,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0731,
      "step": 41021
    },
    {
      "epoch": 9.765339522704279,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 2.4068,
      "step": 41022
    },
    {
      "epoch": 9.765577575432959,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9925,
      "step": 41023
    },
    {
      "epoch": 9.765815628161638,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0553,
      "step": 41024
    },
    {
      "epoch": 9.766053680890318,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.0184,
      "step": 41025
    },
    {
      "epoch": 9.766291733618997,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9743,
      "step": 41026
    },
    {
      "epoch": 9.766529786347675,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 41027
    },
    {
      "epoch": 9.766767839076355,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.9994,
      "step": 41028
    },
    {
      "epoch": 9.767005891805034,
      "grad_norm": 5.90625,
      "learning_rate": 3e-05,
      "loss": 1.8711,
      "step": 41029
    },
    {
      "epoch": 9.767243944533714,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 1.9749,
      "step": 41030
    },
    {
      "epoch": 9.767481997262394,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.233,
      "step": 41031
    },
    {
      "epoch": 9.767720049991073,
      "grad_norm": 5.71875,
      "learning_rate": 3e-05,
      "loss": 1.8256,
      "step": 41032
    },
    {
      "epoch": 9.767958102719753,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.1248,
      "step": 41033
    },
    {
      "epoch": 9.768196155448432,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.019,
      "step": 41034
    },
    {
      "epoch": 9.768434208177112,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1592,
      "step": 41035
    },
    {
      "epoch": 9.76867226090579,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.0139,
      "step": 41036
    },
    {
      "epoch": 9.76891031363447,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.0052,
      "step": 41037
    },
    {
      "epoch": 9.769148366363149,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2081,
      "step": 41038
    },
    {
      "epoch": 9.769386419091829,
      "grad_norm": 5.9375,
      "learning_rate": 3e-05,
      "loss": 1.8265,
      "step": 41039
    },
    {
      "epoch": 9.769624471820508,
      "grad_norm": 6.125,
      "learning_rate": 3e-05,
      "loss": 2.3148,
      "step": 41040
    },
    {
      "epoch": 9.769862524549188,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.3478,
      "step": 41041
    },
    {
      "epoch": 9.770100577277868,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8742,
      "step": 41042
    },
    {
      "epoch": 9.770338630006547,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1197,
      "step": 41043
    },
    {
      "epoch": 9.770576682735225,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.9989,
      "step": 41044
    },
    {
      "epoch": 9.770814735463905,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.9467,
      "step": 41045
    },
    {
      "epoch": 9.771052788192584,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.9374,
      "step": 41046
    },
    {
      "epoch": 9.771290840921264,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1705,
      "step": 41047
    },
    {
      "epoch": 9.771528893649943,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9141,
      "step": 41048
    },
    {
      "epoch": 9.771766946378623,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0067,
      "step": 41049
    },
    {
      "epoch": 9.772004999107303,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.2349,
      "step": 41050
    },
    {
      "epoch": 9.772243051835982,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1631,
      "step": 41051
    },
    {
      "epoch": 9.772481104564662,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9788,
      "step": 41052
    },
    {
      "epoch": 9.772719157293341,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3809,
      "step": 41053
    },
    {
      "epoch": 9.77295721002202,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2445,
      "step": 41054
    },
    {
      "epoch": 9.773195262750699,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.9069,
      "step": 41055
    },
    {
      "epoch": 9.773433315479378,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0903,
      "step": 41056
    },
    {
      "epoch": 9.773671368208058,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 41057
    },
    {
      "epoch": 9.773909420936738,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.1152,
      "step": 41058
    },
    {
      "epoch": 9.774147473665417,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 41059
    },
    {
      "epoch": 9.774385526394097,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9579,
      "step": 41060
    },
    {
      "epoch": 9.774623579122776,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 41061
    },
    {
      "epoch": 9.774861631851454,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8549,
      "step": 41062
    },
    {
      "epoch": 9.775099684580134,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.254,
      "step": 41063
    },
    {
      "epoch": 9.775337737308814,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 1.9755,
      "step": 41064
    },
    {
      "epoch": 9.775575790037493,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9856,
      "step": 41065
    },
    {
      "epoch": 9.775813842766173,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 2.1363,
      "step": 41066
    },
    {
      "epoch": 9.776051895494852,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1146,
      "step": 41067
    },
    {
      "epoch": 9.776289948223532,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2584,
      "step": 41068
    },
    {
      "epoch": 9.776528000952212,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.3254,
      "step": 41069
    },
    {
      "epoch": 9.776766053680891,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2824,
      "step": 41070
    },
    {
      "epoch": 9.777004106409569,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.8375,
      "step": 41071
    },
    {
      "epoch": 9.777242159138249,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.996,
      "step": 41072
    },
    {
      "epoch": 9.777480211866928,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2384,
      "step": 41073
    },
    {
      "epoch": 9.777718264595608,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1846,
      "step": 41074
    },
    {
      "epoch": 9.777956317324287,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.9658,
      "step": 41075
    },
    {
      "epoch": 9.778194370052967,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8922,
      "step": 41076
    },
    {
      "epoch": 9.778432422781647,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 41077
    },
    {
      "epoch": 9.778670475510326,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8789,
      "step": 41078
    },
    {
      "epoch": 9.778908528239004,
      "grad_norm": 2.96875,
      "learning_rate": 3e-05,
      "loss": 1.9527,
      "step": 41079
    },
    {
      "epoch": 9.779146580967684,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.1529,
      "step": 41080
    },
    {
      "epoch": 9.779384633696363,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 1.8904,
      "step": 41081
    },
    {
      "epoch": 9.779622686425043,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.9651,
      "step": 41082
    },
    {
      "epoch": 9.779860739153722,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.3087,
      "step": 41083
    },
    {
      "epoch": 9.780098791882402,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 1.8065,
      "step": 41084
    },
    {
      "epoch": 9.780336844611082,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8277,
      "step": 41085
    },
    {
      "epoch": 9.780574897339761,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.4288,
      "step": 41086
    },
    {
      "epoch": 9.78081295006844,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.8485,
      "step": 41087
    },
    {
      "epoch": 9.78105100279712,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 1.9916,
      "step": 41088
    },
    {
      "epoch": 9.781289055525798,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0152,
      "step": 41089
    },
    {
      "epoch": 9.781527108254478,
      "grad_norm": 5.15625,
      "learning_rate": 3e-05,
      "loss": 1.9078,
      "step": 41090
    },
    {
      "epoch": 9.781765160983158,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.0532,
      "step": 41091
    },
    {
      "epoch": 9.782003213711837,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8158,
      "step": 41092
    },
    {
      "epoch": 9.782241266440517,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.9987,
      "step": 41093
    },
    {
      "epoch": 9.782479319169196,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.4158,
      "step": 41094
    },
    {
      "epoch": 9.782717371897876,
      "grad_norm": 5.875,
      "learning_rate": 3e-05,
      "loss": 2.1991,
      "step": 41095
    },
    {
      "epoch": 9.782955424626556,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 1.8128,
      "step": 41096
    },
    {
      "epoch": 9.783193477355233,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.2016,
      "step": 41097
    },
    {
      "epoch": 9.783431530083913,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.8384,
      "step": 41098
    },
    {
      "epoch": 9.783669582812593,
      "grad_norm": 5.09375,
      "learning_rate": 3e-05,
      "loss": 1.8366,
      "step": 41099
    },
    {
      "epoch": 9.783907635541272,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.1935,
      "step": 41100
    },
    {
      "epoch": 9.783907635541272,
      "eval_runtime": 0.5377,
      "eval_samples_per_second": 1.86,
      "eval_steps_per_second": 1.86,
      "step": 41100
    },
    {
      "epoch": 9.784145688269952,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.137,
      "step": 41101
    },
    {
      "epoch": 9.784383740998631,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9997,
      "step": 41102
    },
    {
      "epoch": 9.784621793727311,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.2556,
      "step": 41103
    },
    {
      "epoch": 9.78485984645599,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0025,
      "step": 41104
    },
    {
      "epoch": 9.78509789918467,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.2065,
      "step": 41105
    },
    {
      "epoch": 9.785335951913348,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 41106
    },
    {
      "epoch": 9.785574004642028,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1512,
      "step": 41107
    },
    {
      "epoch": 9.785812057370707,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8335,
      "step": 41108
    },
    {
      "epoch": 9.786050110099387,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1508,
      "step": 41109
    },
    {
      "epoch": 9.786288162828066,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9161,
      "step": 41110
    },
    {
      "epoch": 9.786526215556746,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2898,
      "step": 41111
    },
    {
      "epoch": 9.786764268285426,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 2.2621,
      "step": 41112
    },
    {
      "epoch": 9.787002321014105,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 41113
    },
    {
      "epoch": 9.787240373742783,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0753,
      "step": 41114
    },
    {
      "epoch": 9.787478426471463,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0411,
      "step": 41115
    },
    {
      "epoch": 9.787716479200142,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 2.3903,
      "step": 41116
    },
    {
      "epoch": 9.787954531928822,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 1.9633,
      "step": 41117
    },
    {
      "epoch": 9.788192584657502,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9741,
      "step": 41118
    },
    {
      "epoch": 9.788430637386181,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 1.9496,
      "step": 41119
    },
    {
      "epoch": 9.78866869011486,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.9717,
      "step": 41120
    },
    {
      "epoch": 9.78890674284354,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8226,
      "step": 41121
    },
    {
      "epoch": 9.78914479557222,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.868,
      "step": 41122
    },
    {
      "epoch": 9.7893828483009,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0518,
      "step": 41123
    },
    {
      "epoch": 9.789620901029577,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9426,
      "step": 41124
    },
    {
      "epoch": 9.789858953758257,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3838,
      "step": 41125
    },
    {
      "epoch": 9.790097006486937,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2276,
      "step": 41126
    },
    {
      "epoch": 9.790335059215616,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.1455,
      "step": 41127
    },
    {
      "epoch": 9.790573111944296,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1062,
      "step": 41128
    },
    {
      "epoch": 9.790811164672975,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 1.8465,
      "step": 41129
    },
    {
      "epoch": 9.791049217401655,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.889,
      "step": 41130
    },
    {
      "epoch": 9.791287270130335,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.1287,
      "step": 41131
    },
    {
      "epoch": 9.791525322859012,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.8705,
      "step": 41132
    },
    {
      "epoch": 9.791763375587692,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 1.8907,
      "step": 41133
    },
    {
      "epoch": 9.792001428316372,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 41134
    },
    {
      "epoch": 9.792239481045051,
      "grad_norm": 8.625,
      "learning_rate": 3e-05,
      "loss": 2.0311,
      "step": 41135
    },
    {
      "epoch": 9.79247753377373,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.7983,
      "step": 41136
    },
    {
      "epoch": 9.79271558650241,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.1622,
      "step": 41137
    },
    {
      "epoch": 9.79295363923109,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2512,
      "step": 41138
    },
    {
      "epoch": 9.79319169195977,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 1.8913,
      "step": 41139
    },
    {
      "epoch": 9.79342974468845,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.5786,
      "step": 41140
    },
    {
      "epoch": 9.793667797417127,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 1.9275,
      "step": 41141
    },
    {
      "epoch": 9.793905850145807,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1913,
      "step": 41142
    },
    {
      "epoch": 9.794143902874486,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0648,
      "step": 41143
    },
    {
      "epoch": 9.794381955603166,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0217,
      "step": 41144
    },
    {
      "epoch": 9.794620008331846,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1541,
      "step": 41145
    },
    {
      "epoch": 9.794858061060525,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.3198,
      "step": 41146
    },
    {
      "epoch": 9.795096113789205,
      "grad_norm": 3.96875,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 41147
    },
    {
      "epoch": 9.795334166517884,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.0371,
      "step": 41148
    },
    {
      "epoch": 9.795572219246562,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.9288,
      "step": 41149
    },
    {
      "epoch": 9.795810271975242,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8454,
      "step": 41150
    },
    {
      "epoch": 9.796048324703921,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.1976,
      "step": 41151
    },
    {
      "epoch": 9.796286377432601,
      "grad_norm": 6.46875,
      "learning_rate": 3e-05,
      "loss": 1.8627,
      "step": 41152
    },
    {
      "epoch": 9.79652443016128,
      "grad_norm": 5.53125,
      "learning_rate": 3e-05,
      "loss": 2.2828,
      "step": 41153
    },
    {
      "epoch": 9.79676248288996,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.068,
      "step": 41154
    },
    {
      "epoch": 9.79700053561864,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 1.8647,
      "step": 41155
    },
    {
      "epoch": 9.79723858834732,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9456,
      "step": 41156
    },
    {
      "epoch": 9.797476641075999,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.125,
      "step": 41157
    },
    {
      "epoch": 9.797714693804679,
      "grad_norm": 5.3125,
      "learning_rate": 3e-05,
      "loss": 1.9365,
      "step": 41158
    },
    {
      "epoch": 9.797952746533356,
      "grad_norm": 5.6875,
      "learning_rate": 3e-05,
      "loss": 1.9524,
      "step": 41159
    },
    {
      "epoch": 9.798190799262036,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.0542,
      "step": 41160
    },
    {
      "epoch": 9.798428851990716,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0254,
      "step": 41161
    },
    {
      "epoch": 9.798666904719395,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.2797,
      "step": 41162
    },
    {
      "epoch": 9.798904957448075,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2641,
      "step": 41163
    },
    {
      "epoch": 9.799143010176754,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.9718,
      "step": 41164
    },
    {
      "epoch": 9.799381062905434,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 2.2681,
      "step": 41165
    },
    {
      "epoch": 9.799619115634114,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 41166
    },
    {
      "epoch": 9.799857168362792,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.0524,
      "step": 41167
    },
    {
      "epoch": 9.800095221091471,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9785,
      "step": 41168
    },
    {
      "epoch": 9.80033327382015,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0761,
      "step": 41169
    },
    {
      "epoch": 9.80057132654883,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.0978,
      "step": 41170
    },
    {
      "epoch": 9.80080937927751,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.249,
      "step": 41171
    },
    {
      "epoch": 9.80104743200619,
      "grad_norm": 5.4375,
      "learning_rate": 3e-05,
      "loss": 1.8606,
      "step": 41172
    },
    {
      "epoch": 9.80128548473487,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 2.2018,
      "step": 41173
    },
    {
      "epoch": 9.801523537463549,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9993,
      "step": 41174
    },
    {
      "epoch": 9.801761590192228,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.8577,
      "step": 41175
    },
    {
      "epoch": 9.801999642920906,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1068,
      "step": 41176
    },
    {
      "epoch": 9.802237695649586,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 1.8211,
      "step": 41177
    },
    {
      "epoch": 9.802475748378265,
      "grad_norm": 3.046875,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 41178
    },
    {
      "epoch": 9.802713801106945,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9037,
      "step": 41179
    },
    {
      "epoch": 9.802951853835625,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.7795,
      "step": 41180
    },
    {
      "epoch": 9.803189906564304,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.4149,
      "step": 41181
    },
    {
      "epoch": 9.803427959292984,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9991,
      "step": 41182
    },
    {
      "epoch": 9.803666012021663,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.9259,
      "step": 41183
    },
    {
      "epoch": 9.803904064750341,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.4134,
      "step": 41184
    },
    {
      "epoch": 9.80414211747902,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 2.0466,
      "step": 41185
    },
    {
      "epoch": 9.8043801702077,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0592,
      "step": 41186
    },
    {
      "epoch": 9.80461822293638,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 2.2669,
      "step": 41187
    },
    {
      "epoch": 9.80485627566506,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.9542,
      "step": 41188
    },
    {
      "epoch": 9.80509432839374,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.1358,
      "step": 41189
    },
    {
      "epoch": 9.805332381122419,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 41190
    },
    {
      "epoch": 9.805570433851098,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.0682,
      "step": 41191
    },
    {
      "epoch": 9.805808486579778,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.9618,
      "step": 41192
    },
    {
      "epoch": 9.806046539308458,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.9706,
      "step": 41193
    },
    {
      "epoch": 9.806284592037136,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1721,
      "step": 41194
    },
    {
      "epoch": 9.806522644765815,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.1151,
      "step": 41195
    },
    {
      "epoch": 9.806760697494495,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.3911,
      "step": 41196
    },
    {
      "epoch": 9.806998750223174,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 41197
    },
    {
      "epoch": 9.807236802951854,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 1.9859,
      "step": 41198
    },
    {
      "epoch": 9.807474855680534,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9441,
      "step": 41199
    },
    {
      "epoch": 9.807712908409213,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0815,
      "step": 41200
    },
    {
      "epoch": 9.807712908409213,
      "eval_runtime": 0.5334,
      "eval_samples_per_second": 1.875,
      "eval_steps_per_second": 1.875,
      "step": 41200
    },
    {
      "epoch": 9.807950961137893,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 1.8545,
      "step": 41201
    },
    {
      "epoch": 9.80818901386657,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0326,
      "step": 41202
    },
    {
      "epoch": 9.80842706659525,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2667,
      "step": 41203
    },
    {
      "epoch": 9.80866511932393,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9021,
      "step": 41204
    },
    {
      "epoch": 9.80890317205261,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.8022,
      "step": 41205
    },
    {
      "epoch": 9.809141224781289,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.3233,
      "step": 41206
    },
    {
      "epoch": 9.809379277509969,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1792,
      "step": 41207
    },
    {
      "epoch": 9.809617330238648,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1318,
      "step": 41208
    },
    {
      "epoch": 9.809855382967328,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.4833,
      "step": 41209
    },
    {
      "epoch": 9.810093435696007,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.7196,
      "step": 41210
    },
    {
      "epoch": 9.810331488424685,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.0327,
      "step": 41211
    },
    {
      "epoch": 9.810569541153365,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.7247,
      "step": 41212
    },
    {
      "epoch": 9.810807593882044,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.026,
      "step": 41213
    },
    {
      "epoch": 9.811045646610724,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3883,
      "step": 41214
    },
    {
      "epoch": 9.811283699339404,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.108,
      "step": 41215
    },
    {
      "epoch": 9.811521752068083,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0907,
      "step": 41216
    },
    {
      "epoch": 9.811759804796763,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 1.921,
      "step": 41217
    },
    {
      "epoch": 9.811997857525443,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9955,
      "step": 41218
    },
    {
      "epoch": 9.812235910254122,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9903,
      "step": 41219
    },
    {
      "epoch": 9.8124739629828,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 1.8911,
      "step": 41220
    },
    {
      "epoch": 9.81271201571148,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 41221
    },
    {
      "epoch": 9.81295006844016,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1265,
      "step": 41222
    },
    {
      "epoch": 9.813188121168839,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.1326,
      "step": 41223
    },
    {
      "epoch": 9.813426173897518,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.8969,
      "step": 41224
    },
    {
      "epoch": 9.813664226626198,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 1.9787,
      "step": 41225
    },
    {
      "epoch": 9.813902279354878,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1112,
      "step": 41226
    },
    {
      "epoch": 9.814140332083557,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.3441,
      "step": 41227
    },
    {
      "epoch": 9.814378384812237,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.221,
      "step": 41228
    },
    {
      "epoch": 9.814616437540915,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.944,
      "step": 41229
    },
    {
      "epoch": 9.814854490269594,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0332,
      "step": 41230
    },
    {
      "epoch": 9.815092542998274,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 1.8679,
      "step": 41231
    },
    {
      "epoch": 9.815330595726953,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.975,
      "step": 41232
    },
    {
      "epoch": 9.815568648455633,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1733,
      "step": 41233
    },
    {
      "epoch": 9.815806701184313,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.8771,
      "step": 41234
    },
    {
      "epoch": 9.816044753912992,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8005,
      "step": 41235
    },
    {
      "epoch": 9.816282806641672,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0202,
      "step": 41236
    },
    {
      "epoch": 9.81652085937035,
      "grad_norm": 5.59375,
      "learning_rate": 3e-05,
      "loss": 1.9575,
      "step": 41237
    },
    {
      "epoch": 9.81675891209903,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.0287,
      "step": 41238
    },
    {
      "epoch": 9.816996964827709,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0919,
      "step": 41239
    },
    {
      "epoch": 9.817235017556388,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.8686,
      "step": 41240
    },
    {
      "epoch": 9.817473070285068,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.771,
      "step": 41241
    },
    {
      "epoch": 9.817711123013748,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.2483,
      "step": 41242
    },
    {
      "epoch": 9.817949175742427,
      "grad_norm": 5.375,
      "learning_rate": 3e-05,
      "loss": 1.9252,
      "step": 41243
    },
    {
      "epoch": 9.818187228471107,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.0015,
      "step": 41244
    },
    {
      "epoch": 9.818425281199787,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.955,
      "step": 41245
    },
    {
      "epoch": 9.818663333928464,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.9022,
      "step": 41246
    },
    {
      "epoch": 9.818901386657144,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.1918,
      "step": 41247
    },
    {
      "epoch": 9.819139439385824,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1182,
      "step": 41248
    },
    {
      "epoch": 9.819377492114503,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0063,
      "step": 41249
    },
    {
      "epoch": 9.819615544843183,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.7044,
      "step": 41250
    },
    {
      "epoch": 9.819853597571862,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1273,
      "step": 41251
    },
    {
      "epoch": 9.820091650300542,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.8862,
      "step": 41252
    },
    {
      "epoch": 9.820329703029222,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.1309,
      "step": 41253
    },
    {
      "epoch": 9.820567755757901,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.0526,
      "step": 41254
    },
    {
      "epoch": 9.820805808486579,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.9746,
      "step": 41255
    },
    {
      "epoch": 9.821043861215259,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8529,
      "step": 41256
    },
    {
      "epoch": 9.821281913943938,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 1.9769,
      "step": 41257
    },
    {
      "epoch": 9.821519966672618,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.2142,
      "step": 41258
    },
    {
      "epoch": 9.821758019401297,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 2.0486,
      "step": 41259
    },
    {
      "epoch": 9.821996072129977,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.7139,
      "step": 41260
    },
    {
      "epoch": 9.822234124858657,
      "grad_norm": 4.8125,
      "learning_rate": 3e-05,
      "loss": 2.3989,
      "step": 41261
    },
    {
      "epoch": 9.822472177587336,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.3049,
      "step": 41262
    },
    {
      "epoch": 9.822710230316016,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.8032,
      "step": 41263
    },
    {
      "epoch": 9.822948283044694,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1407,
      "step": 41264
    },
    {
      "epoch": 9.823186335773373,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0427,
      "step": 41265
    },
    {
      "epoch": 9.823424388502053,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.074,
      "step": 41266
    },
    {
      "epoch": 9.823662441230733,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.2541,
      "step": 41267
    },
    {
      "epoch": 9.823900493959412,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0221,
      "step": 41268
    },
    {
      "epoch": 9.824138546688092,
      "grad_norm": 4.875,
      "learning_rate": 3e-05,
      "loss": 1.9481,
      "step": 41269
    },
    {
      "epoch": 9.824376599416771,
      "grad_norm": 3.578125,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 41270
    },
    {
      "epoch": 9.824614652145451,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0204,
      "step": 41271
    },
    {
      "epoch": 9.824852704874129,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1232,
      "step": 41272
    },
    {
      "epoch": 9.825090757602808,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1234,
      "step": 41273
    },
    {
      "epoch": 9.825328810331488,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 1.986,
      "step": 41274
    },
    {
      "epoch": 9.825566863060168,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8268,
      "step": 41275
    },
    {
      "epoch": 9.825804915788847,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.0932,
      "step": 41276
    },
    {
      "epoch": 9.826042968517527,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9961,
      "step": 41277
    },
    {
      "epoch": 9.826281021246206,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 1.9233,
      "step": 41278
    },
    {
      "epoch": 9.826519073974886,
      "grad_norm": 4.0,
      "learning_rate": 3e-05,
      "loss": 1.8718,
      "step": 41279
    },
    {
      "epoch": 9.826757126703566,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.3348,
      "step": 41280
    },
    {
      "epoch": 9.826995179432243,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.4383,
      "step": 41281
    },
    {
      "epoch": 9.827233232160923,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.1098,
      "step": 41282
    },
    {
      "epoch": 9.827471284889603,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0333,
      "step": 41283
    },
    {
      "epoch": 9.827709337618282,
      "grad_norm": 3.328125,
      "learning_rate": 3e-05,
      "loss": 1.9488,
      "step": 41284
    },
    {
      "epoch": 9.827947390346962,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.7606,
      "step": 41285
    },
    {
      "epoch": 9.828185443075641,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.139,
      "step": 41286
    },
    {
      "epoch": 9.828423495804321,
      "grad_norm": 3.296875,
      "learning_rate": 3e-05,
      "loss": 2.0415,
      "step": 41287
    },
    {
      "epoch": 9.828661548533,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9332,
      "step": 41288
    },
    {
      "epoch": 9.82889960126168,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.9866,
      "step": 41289
    },
    {
      "epoch": 9.829137653990358,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2335,
      "step": 41290
    },
    {
      "epoch": 9.829375706719038,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1433,
      "step": 41291
    },
    {
      "epoch": 9.829613759447717,
      "grad_norm": 3.453125,
      "learning_rate": 3e-05,
      "loss": 2.0967,
      "step": 41292
    },
    {
      "epoch": 9.829851812176397,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9946,
      "step": 41293
    },
    {
      "epoch": 9.830089864905077,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.0671,
      "step": 41294
    },
    {
      "epoch": 9.830327917633756,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.0604,
      "step": 41295
    },
    {
      "epoch": 9.830565970362436,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.8892,
      "step": 41296
    },
    {
      "epoch": 9.830804023091115,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 1.9825,
      "step": 41297
    },
    {
      "epoch": 9.831042075819795,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.0465,
      "step": 41298
    },
    {
      "epoch": 9.831280128548473,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.359,
      "step": 41299
    },
    {
      "epoch": 9.831518181277152,
      "grad_norm": 5.625,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 41300
    },
    {
      "epoch": 9.831518181277152,
      "eval_runtime": 0.5177,
      "eval_samples_per_second": 1.932,
      "eval_steps_per_second": 1.932,
      "step": 41300
    },
    {
      "epoch": 9.831756234005832,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 2.1612,
      "step": 41301
    },
    {
      "epoch": 9.831994286734512,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.0719,
      "step": 41302
    },
    {
      "epoch": 9.832232339463191,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 1.8761,
      "step": 41303
    },
    {
      "epoch": 9.83247039219187,
      "grad_norm": 3.390625,
      "learning_rate": 3e-05,
      "loss": 1.7488,
      "step": 41304
    },
    {
      "epoch": 9.83270844492055,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 1.9164,
      "step": 41305
    },
    {
      "epoch": 9.83294649764923,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.0688,
      "step": 41306
    },
    {
      "epoch": 9.833184550377908,
      "grad_norm": 4.6875,
      "learning_rate": 3e-05,
      "loss": 2.1743,
      "step": 41307
    },
    {
      "epoch": 9.833422603106587,
      "grad_norm": 4.75,
      "learning_rate": 3e-05,
      "loss": 1.8397,
      "step": 41308
    },
    {
      "epoch": 9.833660655835267,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.1708,
      "step": 41309
    },
    {
      "epoch": 9.833898708563947,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.0506,
      "step": 41310
    },
    {
      "epoch": 9.834136761292626,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 1.9635,
      "step": 41311
    },
    {
      "epoch": 9.834374814021306,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 1.6791,
      "step": 41312
    },
    {
      "epoch": 9.834612866749985,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0933,
      "step": 41313
    },
    {
      "epoch": 9.834850919478665,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.0687,
      "step": 41314
    },
    {
      "epoch": 9.835088972207345,
      "grad_norm": 3.125,
      "learning_rate": 3e-05,
      "loss": 1.9879,
      "step": 41315
    },
    {
      "epoch": 9.835327024936023,
      "grad_norm": 4.5625,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 41316
    },
    {
      "epoch": 9.835565077664702,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.1828,
      "step": 41317
    },
    {
      "epoch": 9.835803130393382,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.1519,
      "step": 41318
    },
    {
      "epoch": 9.836041183122061,
      "grad_norm": 3.21875,
      "learning_rate": 3e-05,
      "loss": 2.0463,
      "step": 41319
    },
    {
      "epoch": 9.836279235850741,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1505,
      "step": 41320
    },
    {
      "epoch": 9.83651728857942,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.1119,
      "step": 41321
    },
    {
      "epoch": 9.8367553413081,
      "grad_norm": 2.984375,
      "learning_rate": 3e-05,
      "loss": 1.9229,
      "step": 41322
    },
    {
      "epoch": 9.83699339403678,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0473,
      "step": 41323
    },
    {
      "epoch": 9.83723144676546,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 2.1873,
      "step": 41324
    },
    {
      "epoch": 9.837469499494137,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0861,
      "step": 41325
    },
    {
      "epoch": 9.837707552222817,
      "grad_norm": 5.21875,
      "learning_rate": 3e-05,
      "loss": 2.1798,
      "step": 41326
    },
    {
      "epoch": 9.837945604951496,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9466,
      "step": 41327
    },
    {
      "epoch": 9.838183657680176,
      "grad_norm": 3.90625,
      "learning_rate": 3e-05,
      "loss": 1.8958,
      "step": 41328
    },
    {
      "epoch": 9.838421710408856,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 1.7889,
      "step": 41329
    },
    {
      "epoch": 9.838659763137535,
      "grad_norm": 4.40625,
      "learning_rate": 3e-05,
      "loss": 2.025,
      "step": 41330
    },
    {
      "epoch": 9.838897815866215,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0275,
      "step": 41331
    },
    {
      "epoch": 9.839135868594894,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.8531,
      "step": 41332
    },
    {
      "epoch": 9.839373921323574,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.12,
      "step": 41333
    },
    {
      "epoch": 9.839611974052252,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1865,
      "step": 41334
    },
    {
      "epoch": 9.839850026780931,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.0208,
      "step": 41335
    },
    {
      "epoch": 9.840088079509611,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.035,
      "step": 41336
    },
    {
      "epoch": 9.84032613223829,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8558,
      "step": 41337
    },
    {
      "epoch": 9.84056418496697,
      "grad_norm": 3.40625,
      "learning_rate": 3e-05,
      "loss": 1.8983,
      "step": 41338
    },
    {
      "epoch": 9.84080223769565,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0845,
      "step": 41339
    },
    {
      "epoch": 9.84104029042433,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0555,
      "step": 41340
    },
    {
      "epoch": 9.841278343153009,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.1019,
      "step": 41341
    },
    {
      "epoch": 9.841516395881687,
      "grad_norm": 5.125,
      "learning_rate": 3e-05,
      "loss": 2.0977,
      "step": 41342
    },
    {
      "epoch": 9.841754448610367,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0329,
      "step": 41343
    },
    {
      "epoch": 9.841992501339046,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 2.1106,
      "step": 41344
    },
    {
      "epoch": 9.842230554067726,
      "grad_norm": 4.90625,
      "learning_rate": 3e-05,
      "loss": 2.1134,
      "step": 41345
    },
    {
      "epoch": 9.842468606796405,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.127,
      "step": 41346
    },
    {
      "epoch": 9.842706659525085,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.852,
      "step": 41347
    },
    {
      "epoch": 9.842944712253765,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0001,
      "step": 41348
    },
    {
      "epoch": 9.843182764982444,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.0116,
      "step": 41349
    },
    {
      "epoch": 9.843420817711124,
      "grad_norm": 4.71875,
      "learning_rate": 3e-05,
      "loss": 2.2986,
      "step": 41350
    },
    {
      "epoch": 9.843658870439802,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.2679,
      "step": 41351
    },
    {
      "epoch": 9.843896923168481,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.018,
      "step": 41352
    },
    {
      "epoch": 9.84413497589716,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0935,
      "step": 41353
    },
    {
      "epoch": 9.84437302862584,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.0393,
      "step": 41354
    },
    {
      "epoch": 9.84461108135452,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.2952,
      "step": 41355
    },
    {
      "epoch": 9.8448491340832,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.2148,
      "step": 41356
    },
    {
      "epoch": 9.84508718681188,
      "grad_norm": 3.609375,
      "learning_rate": 3e-05,
      "loss": 1.9034,
      "step": 41357
    },
    {
      "epoch": 9.845325239540559,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 41358
    },
    {
      "epoch": 9.845563292269238,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 1.6827,
      "step": 41359
    },
    {
      "epoch": 9.845801344997916,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.95,
      "step": 41360
    },
    {
      "epoch": 9.846039397726596,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1118,
      "step": 41361
    },
    {
      "epoch": 9.846277450455275,
      "grad_norm": 3.796875,
      "learning_rate": 3e-05,
      "loss": 2.0099,
      "step": 41362
    },
    {
      "epoch": 9.846515503183955,
      "grad_norm": 3.78125,
      "learning_rate": 3e-05,
      "loss": 1.9529,
      "step": 41363
    },
    {
      "epoch": 9.846753555912635,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 2.0365,
      "step": 41364
    },
    {
      "epoch": 9.846991608641314,
      "grad_norm": 4.3125,
      "learning_rate": 3e-05,
      "loss": 2.1066,
      "step": 41365
    },
    {
      "epoch": 9.847229661369994,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.3232,
      "step": 41366
    },
    {
      "epoch": 9.847467714098673,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 1.9615,
      "step": 41367
    },
    {
      "epoch": 9.847705766827353,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 2.0266,
      "step": 41368
    },
    {
      "epoch": 9.847943819556031,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 2.1398,
      "step": 41369
    },
    {
      "epoch": 9.84818187228471,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2518,
      "step": 41370
    },
    {
      "epoch": 9.84841992501339,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.7962,
      "step": 41371
    },
    {
      "epoch": 9.84865797774207,
      "grad_norm": 3.140625,
      "learning_rate": 3e-05,
      "loss": 2.0736,
      "step": 41372
    },
    {
      "epoch": 9.84889603047075,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0309,
      "step": 41373
    },
    {
      "epoch": 9.849134083199429,
      "grad_norm": 3.1875,
      "learning_rate": 3e-05,
      "loss": 1.7432,
      "step": 41374
    },
    {
      "epoch": 9.849372135928109,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8954,
      "step": 41375
    },
    {
      "epoch": 9.849610188656788,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 1.8747,
      "step": 41376
    },
    {
      "epoch": 9.849848241385466,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.1148,
      "step": 41377
    },
    {
      "epoch": 9.850086294114146,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 1.9611,
      "step": 41378
    },
    {
      "epoch": 9.850324346842825,
      "grad_norm": 3.484375,
      "learning_rate": 3e-05,
      "loss": 2.0897,
      "step": 41379
    },
    {
      "epoch": 9.850562399571505,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 2.0068,
      "step": 41380
    },
    {
      "epoch": 9.850800452300184,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 1.8845,
      "step": 41381
    },
    {
      "epoch": 9.851038505028864,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 2.1071,
      "step": 41382
    },
    {
      "epoch": 9.851276557757544,
      "grad_norm": 5.34375,
      "learning_rate": 3e-05,
      "loss": 2.0775,
      "step": 41383
    },
    {
      "epoch": 9.851514610486223,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0764,
      "step": 41384
    },
    {
      "epoch": 9.851752663214903,
      "grad_norm": 3.203125,
      "learning_rate": 3e-05,
      "loss": 1.7969,
      "step": 41385
    },
    {
      "epoch": 9.851990715943582,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 2.1434,
      "step": 41386
    },
    {
      "epoch": 9.85222876867226,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.164,
      "step": 41387
    },
    {
      "epoch": 9.85246682140094,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 2.1679,
      "step": 41388
    },
    {
      "epoch": 9.85270487412962,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.119,
      "step": 41389
    },
    {
      "epoch": 9.852942926858299,
      "grad_norm": 3.875,
      "learning_rate": 3e-05,
      "loss": 2.1241,
      "step": 41390
    },
    {
      "epoch": 9.853180979586979,
      "grad_norm": 3.59375,
      "learning_rate": 3e-05,
      "loss": 1.9193,
      "step": 41391
    },
    {
      "epoch": 9.853419032315658,
      "grad_norm": 3.671875,
      "learning_rate": 3e-05,
      "loss": 2.1296,
      "step": 41392
    },
    {
      "epoch": 9.853657085044338,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 2.1498,
      "step": 41393
    },
    {
      "epoch": 9.853895137773018,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.1538,
      "step": 41394
    },
    {
      "epoch": 9.854133190501695,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 2.0735,
      "step": 41395
    },
    {
      "epoch": 9.854371243230375,
      "grad_norm": 4.125,
      "learning_rate": 3e-05,
      "loss": 1.9812,
      "step": 41396
    },
    {
      "epoch": 9.854609295959055,
      "grad_norm": 3.375,
      "learning_rate": 3e-05,
      "loss": 2.2431,
      "step": 41397
    },
    {
      "epoch": 9.854847348687734,
      "grad_norm": 3.34375,
      "learning_rate": 3e-05,
      "loss": 1.7933,
      "step": 41398
    },
    {
      "epoch": 9.855085401416414,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.195,
      "step": 41399
    },
    {
      "epoch": 9.855323454145093,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.087,
      "step": 41400
    },
    {
      "epoch": 9.855323454145093,
      "eval_runtime": 0.516,
      "eval_samples_per_second": 1.938,
      "eval_steps_per_second": 1.938,
      "step": 41400
    },
    {
      "epoch": 9.855561506873773,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.8511,
      "step": 41401
    },
    {
      "epoch": 9.855799559602453,
      "grad_norm": 3.5625,
      "learning_rate": 3e-05,
      "loss": 1.8893,
      "step": 41402
    },
    {
      "epoch": 9.856037612331132,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 2.069,
      "step": 41403
    },
    {
      "epoch": 9.85627566505981,
      "grad_norm": 4.1875,
      "learning_rate": 3e-05,
      "loss": 1.8602,
      "step": 41404
    },
    {
      "epoch": 9.85651371778849,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 1.9647,
      "step": 41405
    },
    {
      "epoch": 9.85675177051717,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 2.1773,
      "step": 41406
    },
    {
      "epoch": 9.856989823245849,
      "grad_norm": 3.859375,
      "learning_rate": 3e-05,
      "loss": 1.9446,
      "step": 41407
    },
    {
      "epoch": 9.857227875974528,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 2.1093,
      "step": 41408
    },
    {
      "epoch": 9.857465928703208,
      "grad_norm": 3.28125,
      "learning_rate": 3e-05,
      "loss": 2.1799,
      "step": 41409
    },
    {
      "epoch": 9.857703981431888,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 1.9041,
      "step": 41410
    },
    {
      "epoch": 9.857942034160567,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.0102,
      "step": 41411
    },
    {
      "epoch": 9.858180086889245,
      "grad_norm": 3.625,
      "learning_rate": 3e-05,
      "loss": 2.0733,
      "step": 41412
    },
    {
      "epoch": 9.858418139617925,
      "grad_norm": 3.265625,
      "learning_rate": 3e-05,
      "loss": 1.9387,
      "step": 41413
    },
    {
      "epoch": 9.858656192346604,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 1.6673,
      "step": 41414
    },
    {
      "epoch": 9.858894245075284,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1848,
      "step": 41415
    },
    {
      "epoch": 9.859132297803963,
      "grad_norm": 3.421875,
      "learning_rate": 3e-05,
      "loss": 2.0929,
      "step": 41416
    },
    {
      "epoch": 9.859370350532643,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 2.1209,
      "step": 41417
    },
    {
      "epoch": 9.859608403261323,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1836,
      "step": 41418
    },
    {
      "epoch": 9.859846455990002,
      "grad_norm": 3.75,
      "learning_rate": 3e-05,
      "loss": 1.8224,
      "step": 41419
    },
    {
      "epoch": 9.860084508718682,
      "grad_norm": 4.375,
      "learning_rate": 3e-05,
      "loss": 2.2175,
      "step": 41420
    },
    {
      "epoch": 9.860322561447362,
      "grad_norm": 3.953125,
      "learning_rate": 3e-05,
      "loss": 2.0168,
      "step": 41421
    },
    {
      "epoch": 9.86056061417604,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0349,
      "step": 41422
    },
    {
      "epoch": 9.860798666904719,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 2.0335,
      "step": 41423
    },
    {
      "epoch": 9.861036719633399,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0717,
      "step": 41424
    },
    {
      "epoch": 9.861274772362078,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.0406,
      "step": 41425
    },
    {
      "epoch": 9.861512825090758,
      "grad_norm": 3.078125,
      "learning_rate": 3e-05,
      "loss": 1.8067,
      "step": 41426
    },
    {
      "epoch": 9.861750877819437,
      "grad_norm": 3.640625,
      "learning_rate": 3e-05,
      "loss": 1.8975,
      "step": 41427
    },
    {
      "epoch": 9.861988930548117,
      "grad_norm": 3.828125,
      "learning_rate": 3e-05,
      "loss": 2.1814,
      "step": 41428
    },
    {
      "epoch": 9.862226983276797,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.2712,
      "step": 41429
    },
    {
      "epoch": 9.862465036005474,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 1.9975,
      "step": 41430
    },
    {
      "epoch": 9.862703088734154,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.2567,
      "step": 41431
    },
    {
      "epoch": 9.862941141462834,
      "grad_norm": 5.46875,
      "learning_rate": 3e-05,
      "loss": 2.0515,
      "step": 41432
    },
    {
      "epoch": 9.863179194191513,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 1.8886,
      "step": 41433
    },
    {
      "epoch": 9.863417246920193,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.0722,
      "step": 41434
    },
    {
      "epoch": 9.863655299648872,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 1.8939,
      "step": 41435
    },
    {
      "epoch": 9.863893352377552,
      "grad_norm": 3.9375,
      "learning_rate": 3e-05,
      "loss": 2.1211,
      "step": 41436
    },
    {
      "epoch": 9.864131405106232,
      "grad_norm": 3.71875,
      "learning_rate": 3e-05,
      "loss": 2.21,
      "step": 41437
    },
    {
      "epoch": 9.864369457834911,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0334,
      "step": 41438
    },
    {
      "epoch": 9.864607510563589,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0215,
      "step": 41439
    },
    {
      "epoch": 9.864845563292269,
      "grad_norm": 3.65625,
      "learning_rate": 3e-05,
      "loss": 1.9939,
      "step": 41440
    },
    {
      "epoch": 9.865083616020948,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 2.2002,
      "step": 41441
    },
    {
      "epoch": 9.865321668749628,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.366,
      "step": 41442
    },
    {
      "epoch": 9.865559721478308,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 1.9771,
      "step": 41443
    },
    {
      "epoch": 9.865797774206987,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 2.4276,
      "step": 41444
    },
    {
      "epoch": 9.866035826935667,
      "grad_norm": 3.6875,
      "learning_rate": 3e-05,
      "loss": 2.0529,
      "step": 41445
    },
    {
      "epoch": 9.866273879664346,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.7985,
      "step": 41446
    },
    {
      "epoch": 9.866511932393024,
      "grad_norm": 4.96875,
      "learning_rate": 3e-05,
      "loss": 2.0501,
      "step": 41447
    },
    {
      "epoch": 9.866749985121704,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.3897,
      "step": 41448
    },
    {
      "epoch": 9.866988037850383,
      "grad_norm": 5.03125,
      "learning_rate": 3e-05,
      "loss": 1.9624,
      "step": 41449
    },
    {
      "epoch": 9.867226090579063,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.072,
      "step": 41450
    },
    {
      "epoch": 9.867464143307743,
      "grad_norm": 3.4375,
      "learning_rate": 3e-05,
      "loss": 2.0536,
      "step": 41451
    },
    {
      "epoch": 9.867702196036422,
      "grad_norm": 4.84375,
      "learning_rate": 3e-05,
      "loss": 2.2256,
      "step": 41452
    },
    {
      "epoch": 9.867940248765102,
      "grad_norm": 4.5,
      "learning_rate": 3e-05,
      "loss": 2.2843,
      "step": 41453
    },
    {
      "epoch": 9.868178301493781,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0572,
      "step": 41454
    },
    {
      "epoch": 9.868416354222461,
      "grad_norm": 4.53125,
      "learning_rate": 3e-05,
      "loss": 2.1953,
      "step": 41455
    },
    {
      "epoch": 9.86865440695114,
      "grad_norm": 3.15625,
      "learning_rate": 3e-05,
      "loss": 1.7792,
      "step": 41456
    },
    {
      "epoch": 9.868892459679818,
      "grad_norm": 5.0625,
      "learning_rate": 3e-05,
      "loss": 2.4171,
      "step": 41457
    },
    {
      "epoch": 9.869130512408498,
      "grad_norm": 3.3125,
      "learning_rate": 3e-05,
      "loss": 2.1833,
      "step": 41458
    },
    {
      "epoch": 9.869368565137178,
      "grad_norm": 3.109375,
      "learning_rate": 3e-05,
      "loss": 1.8179,
      "step": 41459
    },
    {
      "epoch": 9.869606617865857,
      "grad_norm": 3.890625,
      "learning_rate": 3e-05,
      "loss": 1.8864,
      "step": 41460
    },
    {
      "epoch": 9.869844670594537,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 2.0867,
      "step": 41461
    },
    {
      "epoch": 9.870082723323216,
      "grad_norm": 3.515625,
      "learning_rate": 3e-05,
      "loss": 2.1195,
      "step": 41462
    },
    {
      "epoch": 9.870320776051896,
      "grad_norm": 4.28125,
      "learning_rate": 3e-05,
      "loss": 2.0387,
      "step": 41463
    },
    {
      "epoch": 9.870558828780576,
      "grad_norm": 4.78125,
      "learning_rate": 3e-05,
      "loss": 1.8566,
      "step": 41464
    },
    {
      "epoch": 9.870796881509253,
      "grad_norm": 3.5,
      "learning_rate": 3e-05,
      "loss": 1.9715,
      "step": 41465
    },
    {
      "epoch": 9.871034934237933,
      "grad_norm": 3.8125,
      "learning_rate": 3e-05,
      "loss": 2.2129,
      "step": 41466
    },
    {
      "epoch": 9.871272986966613,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1103,
      "step": 41467
    },
    {
      "epoch": 9.871511039695292,
      "grad_norm": 6.78125,
      "learning_rate": 3e-05,
      "loss": 2.5815,
      "step": 41468
    },
    {
      "epoch": 9.871749092423972,
      "grad_norm": 3.984375,
      "learning_rate": 3e-05,
      "loss": 2.0593,
      "step": 41469
    },
    {
      "epoch": 9.871987145152652,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.0971,
      "step": 41470
    },
    {
      "epoch": 9.872225197881331,
      "grad_norm": 3.546875,
      "learning_rate": 3e-05,
      "loss": 2.1179,
      "step": 41471
    },
    {
      "epoch": 9.87246325061001,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.2143,
      "step": 41472
    },
    {
      "epoch": 9.87270130333869,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.507,
      "step": 41473
    },
    {
      "epoch": 9.872939356067368,
      "grad_norm": 4.0625,
      "learning_rate": 3e-05,
      "loss": 2.0742,
      "step": 41474
    },
    {
      "epoch": 9.873177408796048,
      "grad_norm": 3.703125,
      "learning_rate": 3e-05,
      "loss": 2.0596,
      "step": 41475
    },
    {
      "epoch": 9.873415461524727,
      "grad_norm": 3.53125,
      "learning_rate": 3e-05,
      "loss": 1.8721,
      "step": 41476
    },
    {
      "epoch": 9.873653514253407,
      "grad_norm": 3.734375,
      "learning_rate": 3e-05,
      "loss": 2.0579,
      "step": 41477
    },
    {
      "epoch": 9.873891566982087,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 1.9637,
      "step": 41478
    },
    {
      "epoch": 9.874129619710766,
      "grad_norm": 4.21875,
      "learning_rate": 3e-05,
      "loss": 2.2079,
      "step": 41479
    },
    {
      "epoch": 9.874367672439446,
      "grad_norm": 5.1875,
      "learning_rate": 3e-05,
      "loss": 2.1861,
      "step": 41480
    },
    {
      "epoch": 9.874605725168125,
      "grad_norm": 4.46875,
      "learning_rate": 3e-05,
      "loss": 1.7844,
      "step": 41481
    },
    {
      "epoch": 9.874843777896803,
      "grad_norm": 4.59375,
      "learning_rate": 3e-05,
      "loss": 2.0893,
      "step": 41482
    },
    {
      "epoch": 9.875081830625483,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 2.1379,
      "step": 41483
    },
    {
      "epoch": 9.875319883354162,
      "grad_norm": 4.625,
      "learning_rate": 3e-05,
      "loss": 2.0715,
      "step": 41484
    },
    {
      "epoch": 9.875557936082842,
      "grad_norm": 4.9375,
      "learning_rate": 3e-05,
      "loss": 2.4444,
      "step": 41485
    },
    {
      "epoch": 9.875795988811522,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.245,
      "step": 41486
    },
    {
      "epoch": 9.876034041540201,
      "grad_norm": 5.40625,
      "learning_rate": 3e-05,
      "loss": 2.0246,
      "step": 41487
    },
    {
      "epoch": 9.87627209426888,
      "grad_norm": 3.765625,
      "learning_rate": 3e-05,
      "loss": 1.896,
      "step": 41488
    },
    {
      "epoch": 9.87651014699756,
      "grad_norm": 4.34375,
      "learning_rate": 3e-05,
      "loss": 1.958,
      "step": 41489
    },
    {
      "epoch": 9.87674819972624,
      "grad_norm": 4.25,
      "learning_rate": 3e-05,
      "loss": 2.2168,
      "step": 41490
    },
    {
      "epoch": 9.87698625245492,
      "grad_norm": 4.4375,
      "learning_rate": 3e-05,
      "loss": 1.878,
      "step": 41491
    },
    {
      "epoch": 9.877224305183598,
      "grad_norm": 4.09375,
      "learning_rate": 3e-05,
      "loss": 1.8792,
      "step": 41492
    },
    {
      "epoch": 9.877462357912277,
      "grad_norm": 3.46875,
      "learning_rate": 3e-05,
      "loss": 1.7376,
      "step": 41493
    },
    {
      "epoch": 9.877700410640957,
      "grad_norm": 3.84375,
      "learning_rate": 3e-05,
      "loss": 2.2249,
      "step": 41494
    },
    {
      "epoch": 9.877938463369636,
      "grad_norm": 3.921875,
      "learning_rate": 3e-05,
      "loss": 2.03,
      "step": 41495
    },
    {
      "epoch": 9.878176516098316,
      "grad_norm": 4.15625,
      "learning_rate": 3e-05,
      "loss": 2.0241,
      "step": 41496
    },
    {
      "epoch": 9.878414568826996,
      "grad_norm": 3.171875,
      "learning_rate": 3e-05,
      "loss": 1.8086,
      "step": 41497
    },
    {
      "epoch": 9.878652621555675,
      "grad_norm": 5.5625,
      "learning_rate": 3e-05,
      "loss": 1.9732,
      "step": 41498
    },
    {
      "epoch": 9.878890674284355,
      "grad_norm": 3.359375,
      "learning_rate": 3e-05,
      "loss": 2.1484,
      "step": 41499
    },
    {
      "epoch": 9.879128727013033,
      "grad_norm": 4.03125,
      "learning_rate": 3e-05,
      "loss": 1.8063,
      "step": 41500
    },
    {
      "epoch": 9.879128727013033,
      "eval_runtime": 0.526,
      "eval_samples_per_second": 1.901,
      "eval_steps_per_second": 1.901,
      "step": 41500
    }
  ],
  "logging_steps": 1,
  "max_steps": 42000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.821742941920297e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}