{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 0,
  "global_step": 128,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0078125,
      "grad_norm": 0.61328125,
      "learning_rate": 1e-05,
      "loss": 2.1689,
      "step": 1
    },
    {
      "epoch": 0.015625,
      "grad_norm": 0.58984375,
      "learning_rate": 9.921875e-06,
      "loss": 2.1485,
      "step": 2
    },
    {
      "epoch": 0.0234375,
      "grad_norm": 0.59375,
      "learning_rate": 9.84375e-06,
      "loss": 2.1564,
      "step": 3
    },
    {
      "epoch": 0.03125,
      "grad_norm": 0.5546875,
      "learning_rate": 9.765625e-06,
      "loss": 2.2022,
      "step": 4
    },
    {
      "epoch": 0.0390625,
      "grad_norm": 0.55078125,
      "learning_rate": 9.6875e-06,
      "loss": 2.0902,
      "step": 5
    },
    {
      "epoch": 0.046875,
      "grad_norm": 0.5859375,
      "learning_rate": 9.609375000000001e-06,
      "loss": 2.1296,
      "step": 6
    },
    {
      "epoch": 0.0546875,
      "grad_norm": 0.51171875,
      "learning_rate": 9.531250000000001e-06,
      "loss": 2.1615,
      "step": 7
    },
    {
      "epoch": 0.0625,
      "grad_norm": 0.91015625,
      "learning_rate": 9.453125000000001e-06,
      "loss": 2.2045,
      "step": 8
    },
    {
      "epoch": 0.0703125,
      "grad_norm": 0.474609375,
      "learning_rate": 9.375000000000001e-06,
      "loss": 2.0598,
      "step": 9
    },
    {
      "epoch": 0.078125,
      "grad_norm": 0.47265625,
      "learning_rate": 9.296875e-06,
      "loss": 2.0913,
      "step": 10
    },
    {
      "epoch": 0.0859375,
      "grad_norm": 0.462890625,
      "learning_rate": 9.21875e-06,
      "loss": 2.0043,
      "step": 11
    },
    {
      "epoch": 0.09375,
      "grad_norm": 0.4609375,
      "learning_rate": 9.140625e-06,
      "loss": 1.973,
      "step": 12
    },
    {
      "epoch": 0.1015625,
      "grad_norm": 0.4609375,
      "learning_rate": 9.0625e-06,
      "loss": 1.9907,
      "step": 13
    },
    {
      "epoch": 0.109375,
      "grad_norm": 0.478515625,
      "learning_rate": 8.984375000000002e-06,
      "loss": 2.0225,
      "step": 14
    },
    {
      "epoch": 0.1171875,
      "grad_norm": 0.455078125,
      "learning_rate": 8.906250000000001e-06,
      "loss": 1.868,
      "step": 15
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.48046875,
      "learning_rate": 8.828125000000001e-06,
      "loss": 1.9023,
      "step": 16
    },
    {
      "epoch": 0.1328125,
      "grad_norm": 0.474609375,
      "learning_rate": 8.750000000000001e-06,
      "loss": 1.8461,
      "step": 17
    },
    {
      "epoch": 0.140625,
      "grad_norm": 0.46875,
      "learning_rate": 8.671875e-06,
      "loss": 1.8176,
      "step": 18
    },
    {
      "epoch": 0.1484375,
      "grad_norm": 0.474609375,
      "learning_rate": 8.59375e-06,
      "loss": 1.7922,
      "step": 19
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.4375,
      "learning_rate": 8.515625e-06,
      "loss": 1.7756,
      "step": 20
    },
    {
      "epoch": 0.1640625,
      "grad_norm": 0.421875,
      "learning_rate": 8.4375e-06,
      "loss": 1.7932,
      "step": 21
    },
    {
      "epoch": 0.171875,
      "grad_norm": 0.408203125,
      "learning_rate": 8.359375e-06,
      "loss": 1.702,
      "step": 22
    },
    {
      "epoch": 0.1796875,
      "grad_norm": 0.4375,
      "learning_rate": 8.281250000000001e-06,
      "loss": 1.7802,
      "step": 23
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.392578125,
      "learning_rate": 8.203125000000001e-06,
      "loss": 1.759,
      "step": 24
    },
    {
      "epoch": 0.1953125,
      "grad_norm": 0.357421875,
      "learning_rate": 8.125000000000001e-06,
      "loss": 1.5837,
      "step": 25
    },
    {
      "epoch": 0.203125,
      "grad_norm": 0.357421875,
      "learning_rate": 8.046875e-06,
      "loss": 1.6122,
      "step": 26
    },
    {
      "epoch": 0.2109375,
      "grad_norm": 0.392578125,
      "learning_rate": 7.96875e-06,
      "loss": 1.7369,
      "step": 27
    },
    {
      "epoch": 0.21875,
      "grad_norm": 0.341796875,
      "learning_rate": 7.890625e-06,
      "loss": 1.5856,
      "step": 28
    },
    {
      "epoch": 0.2265625,
      "grad_norm": 0.34375,
      "learning_rate": 7.8125e-06,
      "loss": 1.5203,
      "step": 29
    },
    {
      "epoch": 0.234375,
      "grad_norm": 0.322265625,
      "learning_rate": 7.734375e-06,
      "loss": 1.5575,
      "step": 30
    },
    {
      "epoch": 0.2421875,
      "grad_norm": 0.33203125,
      "learning_rate": 7.656250000000001e-06,
      "loss": 1.6101,
      "step": 31
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.361328125,
      "learning_rate": 7.578125e-06,
      "loss": 1.6346,
      "step": 32
    },
    {
      "epoch": 0.2578125,
      "grad_norm": 0.486328125,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.5728,
      "step": 33
    },
    {
      "epoch": 0.265625,
      "grad_norm": 0.32421875,
      "learning_rate": 7.421875000000001e-06,
      "loss": 1.5461,
      "step": 34
    },
    {
      "epoch": 0.2734375,
      "grad_norm": 0.3046875,
      "learning_rate": 7.343750000000001e-06,
      "loss": 1.4986,
      "step": 35
    },
    {
      "epoch": 0.28125,
      "grad_norm": 0.33203125,
      "learning_rate": 7.265625e-06,
      "loss": 1.5123,
      "step": 36
    },
    {
      "epoch": 0.2890625,
      "grad_norm": 0.28515625,
      "learning_rate": 7.1875e-06,
      "loss": 1.4697,
      "step": 37
    },
    {
      "epoch": 0.296875,
      "grad_norm": 0.2890625,
      "learning_rate": 7.109375000000001e-06,
      "loss": 1.501,
      "step": 38
    },
    {
      "epoch": 0.3046875,
      "grad_norm": 0.296875,
      "learning_rate": 7.031250000000001e-06,
      "loss": 1.4536,
      "step": 39
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.376953125,
      "learning_rate": 6.9531250000000004e-06,
      "loss": 1.5039,
      "step": 40
    },
    {
      "epoch": 0.3203125,
      "grad_norm": 0.28125,
      "learning_rate": 6.875e-06,
      "loss": 1.4613,
      "step": 41
    },
    {
      "epoch": 0.328125,
      "grad_norm": 0.287109375,
      "learning_rate": 6.796875000000001e-06,
      "loss": 1.4792,
      "step": 42
    },
    {
      "epoch": 0.3359375,
      "grad_norm": 0.251953125,
      "learning_rate": 6.718750000000001e-06,
      "loss": 1.4381,
      "step": 43
    },
    {
      "epoch": 0.34375,
      "grad_norm": 0.26171875,
      "learning_rate": 6.6406250000000005e-06,
      "loss": 1.4575,
      "step": 44
    },
    {
      "epoch": 0.3515625,
      "grad_norm": 0.2890625,
      "learning_rate": 6.5625e-06,
      "loss": 1.4558,
      "step": 45
    },
    {
      "epoch": 0.359375,
      "grad_norm": 0.26171875,
      "learning_rate": 6.484375000000001e-06,
      "loss": 1.4537,
      "step": 46
    },
    {
      "epoch": 0.3671875,
      "grad_norm": 0.2421875,
      "learning_rate": 6.406250000000001e-06,
      "loss": 1.3925,
      "step": 47
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.275390625,
      "learning_rate": 6.3281250000000005e-06,
      "loss": 1.5034,
      "step": 48
    },
    {
      "epoch": 0.3828125,
      "grad_norm": 0.27734375,
      "learning_rate": 6.25e-06,
      "loss": 1.4933,
      "step": 49
    },
    {
      "epoch": 0.390625,
      "grad_norm": 0.27734375,
      "learning_rate": 6.171875e-06,
      "loss": 1.4682,
      "step": 50
    },
    {
      "epoch": 0.3984375,
      "grad_norm": 0.2412109375,
      "learning_rate": 6.093750000000001e-06,
      "loss": 1.3798,
      "step": 51
    },
    {
      "epoch": 0.40625,
      "grad_norm": 0.2412109375,
      "learning_rate": 6.0156250000000005e-06,
      "loss": 1.4287,
      "step": 52
    },
    {
      "epoch": 0.4140625,
      "grad_norm": 0.259765625,
      "learning_rate": 5.9375e-06,
      "loss": 1.4116,
      "step": 53
    },
    {
      "epoch": 0.421875,
      "grad_norm": 0.27734375,
      "learning_rate": 5.859375e-06,
      "loss": 1.391,
      "step": 54
    },
    {
      "epoch": 0.4296875,
      "grad_norm": 0.24609375,
      "learning_rate": 5.781250000000001e-06,
      "loss": 1.366,
      "step": 55
    },
    {
      "epoch": 0.4375,
      "grad_norm": 0.23828125,
      "learning_rate": 5.7031250000000006e-06,
      "loss": 1.3947,
      "step": 56
    },
    {
      "epoch": 0.4453125,
      "grad_norm": 0.228515625,
      "learning_rate": 5.625e-06,
      "loss": 1.3593,
      "step": 57
    },
    {
      "epoch": 0.453125,
      "grad_norm": 0.251953125,
      "learning_rate": 5.546875e-06,
      "loss": 1.3858,
      "step": 58
    },
    {
      "epoch": 0.4609375,
      "grad_norm": 0.412109375,
      "learning_rate": 5.468750000000001e-06,
      "loss": 1.2658,
      "step": 59
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.2373046875,
      "learning_rate": 5.390625000000001e-06,
      "loss": 1.3146,
      "step": 60
    },
    {
      "epoch": 0.4765625,
      "grad_norm": 0.2333984375,
      "learning_rate": 5.3125e-06,
      "loss": 1.3277,
      "step": 61
    },
    {
      "epoch": 0.484375,
      "grad_norm": 0.2353515625,
      "learning_rate": 5.234375e-06,
      "loss": 1.3387,
      "step": 62
    },
    {
      "epoch": 0.4921875,
      "grad_norm": 0.236328125,
      "learning_rate": 5.156250000000001e-06,
      "loss": 1.2693,
      "step": 63
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.283203125,
      "learning_rate": 5.078125000000001e-06,
      "loss": 1.4156,
      "step": 64
    },
    {
      "epoch": 0.5078125,
      "grad_norm": 0.22265625,
      "learning_rate": 5e-06,
      "loss": 1.3413,
      "step": 65
    },
    {
      "epoch": 0.515625,
      "grad_norm": 0.27734375,
      "learning_rate": 4.921875e-06,
      "loss": 1.2651,
      "step": 66
    },
    {
      "epoch": 0.5234375,
      "grad_norm": 0.28515625,
      "learning_rate": 4.84375e-06,
      "loss": 1.2676,
      "step": 67
    },
    {
      "epoch": 0.53125,
      "grad_norm": 0.28515625,
      "learning_rate": 4.765625000000001e-06,
      "loss": 1.3619,
      "step": 68
    },
    {
      "epoch": 0.5390625,
      "grad_norm": 0.265625,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 1.293,
      "step": 69
    },
    {
      "epoch": 0.546875,
      "grad_norm": 0.294921875,
      "learning_rate": 4.609375e-06,
      "loss": 1.3701,
      "step": 70
    },
    {
      "epoch": 0.5546875,
      "grad_norm": 0.263671875,
      "learning_rate": 4.53125e-06,
      "loss": 1.3667,
      "step": 71
    },
    {
      "epoch": 0.5625,
      "grad_norm": 0.216796875,
      "learning_rate": 4.453125000000001e-06,
      "loss": 1.312,
      "step": 72
    },
    {
      "epoch": 0.5703125,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 1.2787,
      "step": 73
    },
    {
      "epoch": 0.578125,
      "grad_norm": 0.2021484375,
      "learning_rate": 4.296875e-06,
      "loss": 1.2332,
      "step": 74
    },
    {
      "epoch": 0.5859375,
      "grad_norm": 0.216796875,
      "learning_rate": 4.21875e-06,
      "loss": 1.2995,
      "step": 75
    },
    {
      "epoch": 0.59375,
      "grad_norm": 0.296875,
      "learning_rate": 4.140625000000001e-06,
      "loss": 1.2753,
      "step": 76
    },
    {
      "epoch": 0.6015625,
      "grad_norm": 0.2080078125,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 1.2731,
      "step": 77
    },
    {
      "epoch": 0.609375,
      "grad_norm": 0.25390625,
      "learning_rate": 3.984375e-06,
      "loss": 1.2729,
      "step": 78
    },
    {
      "epoch": 0.6171875,
      "grad_norm": 0.224609375,
      "learning_rate": 3.90625e-06,
      "loss": 1.2847,
      "step": 79
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.2353515625,
      "learning_rate": 3.828125000000001e-06,
      "loss": 1.2964,
      "step": 80
    },
    {
      "epoch": 0.6328125,
      "grad_norm": 0.2177734375,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.3205,
      "step": 81
    },
    {
      "epoch": 0.640625,
      "grad_norm": 0.2119140625,
      "learning_rate": 3.6718750000000003e-06,
      "loss": 1.2755,
      "step": 82
    },
    {
      "epoch": 0.6484375,
      "grad_norm": 0.21875,
      "learning_rate": 3.59375e-06,
      "loss": 1.2444,
      "step": 83
    },
    {
      "epoch": 0.65625,
      "grad_norm": 0.21875,
      "learning_rate": 3.5156250000000003e-06,
      "loss": 1.2853,
      "step": 84
    },
    {
      "epoch": 0.6640625,
      "grad_norm": 0.337890625,
      "learning_rate": 3.4375e-06,
      "loss": 1.3118,
      "step": 85
    },
    {
      "epoch": 0.671875,
      "grad_norm": 0.310546875,
      "learning_rate": 3.3593750000000003e-06,
      "loss": 1.2616,
      "step": 86
    },
    {
      "epoch": 0.6796875,
      "grad_norm": 0.2099609375,
      "learning_rate": 3.28125e-06,
      "loss": 1.2478,
      "step": 87
    },
    {
      "epoch": 0.6875,
      "grad_norm": 0.34765625,
      "learning_rate": 3.2031250000000004e-06,
      "loss": 1.2478,
      "step": 88
    },
    {
      "epoch": 0.6953125,
      "grad_norm": 0.19921875,
      "learning_rate": 3.125e-06,
      "loss": 1.1987,
      "step": 89
    },
    {
      "epoch": 0.703125,
      "grad_norm": 0.224609375,
      "learning_rate": 3.0468750000000004e-06,
      "loss": 1.3421,
      "step": 90
    },
    {
      "epoch": 0.7109375,
      "grad_norm": 0.203125,
      "learning_rate": 2.96875e-06,
      "loss": 1.2491,
      "step": 91
    },
    {
      "epoch": 0.71875,
      "grad_norm": 0.205078125,
      "learning_rate": 2.8906250000000004e-06,
      "loss": 1.2191,
      "step": 92
    },
    {
      "epoch": 0.7265625,
      "grad_norm": 0.212890625,
      "learning_rate": 2.8125e-06,
      "loss": 1.2694,
      "step": 93
    },
    {
      "epoch": 0.734375,
      "grad_norm": 0.283203125,
      "learning_rate": 2.7343750000000004e-06,
      "loss": 1.3663,
      "step": 94
    },
    {
      "epoch": 0.7421875,
      "grad_norm": 0.27734375,
      "learning_rate": 2.65625e-06,
      "loss": 1.29,
      "step": 95
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.240234375,
      "learning_rate": 2.5781250000000004e-06,
      "loss": 1.266,
      "step": 96
    },
    {
      "epoch": 0.7578125,
      "grad_norm": 0.2275390625,
      "learning_rate": 2.5e-06,
      "loss": 1.2643,
      "step": 97
    },
    {
      "epoch": 0.765625,
      "grad_norm": 0.2392578125,
      "learning_rate": 2.421875e-06,
      "loss": 1.2281,
      "step": 98
    },
    {
      "epoch": 0.7734375,
      "grad_norm": 0.1982421875,
      "learning_rate": 2.3437500000000002e-06,
      "loss": 1.1966,
      "step": 99
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.197265625,
      "learning_rate": 2.265625e-06,
      "loss": 1.2174,
      "step": 100
    },
    {
      "epoch": 0.7890625,
      "grad_norm": 0.2001953125,
      "learning_rate": 2.1875000000000002e-06,
      "loss": 1.1622,
      "step": 101
    },
    {
      "epoch": 0.796875,
      "grad_norm": 0.2041015625,
      "learning_rate": 2.109375e-06,
      "loss": 1.2519,
      "step": 102
    },
    {
      "epoch": 0.8046875,
      "grad_norm": 0.2392578125,
      "learning_rate": 2.0312500000000002e-06,
      "loss": 1.3168,
      "step": 103
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.298828125,
      "learning_rate": 1.953125e-06,
      "loss": 1.2386,
      "step": 104
    },
    {
      "epoch": 0.8203125,
      "grad_norm": 0.2421875,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.2855,
      "step": 105
    },
    {
      "epoch": 0.828125,
      "grad_norm": 0.240234375,
      "learning_rate": 1.796875e-06,
      "loss": 1.2156,
      "step": 106
    },
    {
      "epoch": 0.8359375,
      "grad_norm": 0.255859375,
      "learning_rate": 1.71875e-06,
      "loss": 1.2288,
      "step": 107
    },
    {
      "epoch": 0.84375,
      "grad_norm": 0.2255859375,
      "learning_rate": 1.640625e-06,
      "loss": 1.2893,
      "step": 108
    },
    {
      "epoch": 0.8515625,
      "grad_norm": 0.2451171875,
      "learning_rate": 1.5625e-06,
      "loss": 1.2798,
      "step": 109
    },
    {
      "epoch": 0.859375,
      "grad_norm": 0.2451171875,
      "learning_rate": 1.484375e-06,
      "loss": 1.28,
      "step": 110
    },
    {
      "epoch": 0.8671875,
      "grad_norm": 0.208984375,
      "learning_rate": 1.40625e-06,
      "loss": 1.2544,
      "step": 111
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.275390625,
      "learning_rate": 1.328125e-06,
      "loss": 1.2865,
      "step": 112
    },
    {
      "epoch": 0.8828125,
      "grad_norm": 0.2119140625,
      "learning_rate": 1.25e-06,
      "loss": 1.2482,
      "step": 113
    },
    {
      "epoch": 0.890625,
      "grad_norm": 0.197265625,
      "learning_rate": 1.1718750000000001e-06,
      "loss": 1.1816,
      "step": 114
    },
    {
      "epoch": 0.8984375,
      "grad_norm": 0.208984375,
      "learning_rate": 1.0937500000000001e-06,
      "loss": 1.2452,
      "step": 115
    },
    {
      "epoch": 0.90625,
      "grad_norm": 0.2041015625,
      "learning_rate": 1.0156250000000001e-06,
      "loss": 1.2285,
      "step": 116
    },
    {
      "epoch": 0.9140625,
      "grad_norm": 0.2177734375,
      "learning_rate": 9.375000000000001e-07,
      "loss": 1.2178,
      "step": 117
    },
    {
      "epoch": 0.921875,
      "grad_norm": 0.216796875,
      "learning_rate": 8.59375e-07,
      "loss": 1.2501,
      "step": 118
    },
    {
      "epoch": 0.9296875,
      "grad_norm": 0.2080078125,
      "learning_rate": 7.8125e-07,
      "loss": 1.1922,
      "step": 119
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.2138671875,
      "learning_rate": 7.03125e-07,
      "loss": 1.247,
      "step": 120
    },
    {
      "epoch": 0.9453125,
      "grad_norm": 0.2080078125,
      "learning_rate": 6.25e-07,
      "loss": 1.2454,
      "step": 121
    },
    {
      "epoch": 0.953125,
      "grad_norm": 0.2001953125,
      "learning_rate": 5.468750000000001e-07,
      "loss": 1.1875,
      "step": 122
    },
    {
      "epoch": 0.9609375,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.6875000000000006e-07,
      "loss": 1.254,
      "step": 123
    },
    {
      "epoch": 0.96875,
      "grad_norm": 0.2197265625,
      "learning_rate": 3.90625e-07,
      "loss": 1.2924,
      "step": 124
    },
    {
      "epoch": 0.9765625,
      "grad_norm": 0.2119140625,
      "learning_rate": 3.125e-07,
      "loss": 1.2041,
      "step": 125
    },
    {
      "epoch": 0.984375,
      "grad_norm": 0.2392578125,
      "learning_rate": 2.3437500000000003e-07,
      "loss": 1.2446,
      "step": 126
    },
    {
      "epoch": 0.9921875,
      "grad_norm": 0.212890625,
      "learning_rate": 1.5625e-07,
      "loss": 1.2699,
      "step": 127
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2314453125,
      "learning_rate": 7.8125e-08,
      "loss": 1.2642,
      "step": 128
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 128,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.109983433055273e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}