{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 252,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.023809523809523808,
      "grad_norm": 26.742956161499023,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 4.1563,
      "step": 1
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 26.336389541625977,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 4.0633,
      "step": 2
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 28.836008071899414,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 4.3575,
      "step": 3
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 27.459413528442383,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 4.1943,
      "step": 4
    },
    {
      "epoch": 0.11904761904761904,
      "grad_norm": 27.914960861206055,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 4.3336,
      "step": 5
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 27.47397232055664,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 4.2354,
      "step": 6
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 26.212989807128906,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 4.1001,
      "step": 7
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 26.80431365966797,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.1827,
      "step": 8
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 27.641605377197266,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 4.2596,
      "step": 9
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 27.783071517944336,
      "learning_rate": 5.000000000000001e-07,
      "loss": 4.2694,
      "step": 10
    },
    {
      "epoch": 0.2619047619047619,
      "grad_norm": 26.553335189819336,
      "learning_rate": 5.5e-07,
      "loss": 4.0739,
      "step": 11
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 26.77140998840332,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.1049,
      "step": 12
    },
    {
      "epoch": 0.30952380952380953,
      "grad_norm": 25.07087516784668,
      "learning_rate": 6.5e-07,
      "loss": 3.8982,
      "step": 13
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 25.834062576293945,
      "learning_rate": 7.000000000000001e-07,
      "loss": 4.0768,
      "step": 14
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 25.00474739074707,
      "learning_rate": 7.5e-07,
      "loss": 4.0104,
      "step": 15
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 23.692975997924805,
      "learning_rate": 8.000000000000001e-07,
      "loss": 3.8469,
      "step": 16
    },
    {
      "epoch": 0.40476190476190477,
      "grad_norm": 22.197919845581055,
      "learning_rate": 8.500000000000001e-07,
      "loss": 3.7738,
      "step": 17
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 20.92680549621582,
      "learning_rate": 9.000000000000001e-07,
      "loss": 3.6514,
      "step": 18
    },
    {
      "epoch": 0.4523809523809524,
      "grad_norm": 20.251178741455078,
      "learning_rate": 9.500000000000001e-07,
      "loss": 3.6739,
      "step": 19
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 17.55536460876465,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 3.2675,
      "step": 20
    },
    {
      "epoch": 0.5,
      "grad_norm": 17.066797256469727,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 3.4232,
      "step": 21
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 16.1475887298584,
      "learning_rate": 1.1e-06,
      "loss": 3.1988,
      "step": 22
    },
    {
      "epoch": 0.5476190476190477,
      "grad_norm": 15.61026382446289,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 3.1338,
      "step": 23
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 15.409480094909668,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.9836,
      "step": 24
    },
    {
      "epoch": 0.5952380952380952,
      "grad_norm": 15.391901969909668,
      "learning_rate": 1.25e-06,
      "loss": 2.9064,
      "step": 25
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 16.92401885986328,
      "learning_rate": 1.3e-06,
      "loss": 2.906,
      "step": 26
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 17.880958557128906,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 2.702,
      "step": 27
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 18.114517211914062,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.4876,
      "step": 28
    },
    {
      "epoch": 0.6904761904761905,
      "grad_norm": 17.608840942382812,
      "learning_rate": 1.45e-06,
      "loss": 2.2996,
      "step": 29
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 17.055673599243164,
      "learning_rate": 1.5e-06,
      "loss": 2.2709,
      "step": 30
    },
    {
      "epoch": 0.7380952380952381,
      "grad_norm": 14.92151927947998,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.0406,
      "step": 31
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 13.657073020935059,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.8564,
      "step": 32
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 13.274576187133789,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 1.7382,
      "step": 33
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 13.728348731994629,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.6629,
      "step": 34
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 13.521151542663574,
      "learning_rate": 1.75e-06,
      "loss": 1.5301,
      "step": 35
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 13.709525108337402,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.4122,
      "step": 36
    },
    {
      "epoch": 0.8809523809523809,
      "grad_norm": 12.587928771972656,
      "learning_rate": 1.85e-06,
      "loss": 1.2014,
      "step": 37
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 13.494888305664062,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.1793,
      "step": 38
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 13.417922019958496,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 1.0282,
      "step": 39
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 12.754359245300293,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.8525,
      "step": 40
    },
    {
      "epoch": 0.9761904761904762,
      "grad_norm": 12.766407012939453,
      "learning_rate": 2.05e-06,
      "loss": 0.7043,
      "step": 41
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.440421104431152,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.6245,
      "step": 42
    },
    {
      "epoch": 1.0238095238095237,
      "grad_norm": 9.672205924987793,
      "learning_rate": 2.15e-06,
      "loss": 0.4768,
      "step": 43
    },
    {
      "epoch": 1.0476190476190477,
      "grad_norm": 7.8501057624816895,
      "learning_rate": 2.2e-06,
      "loss": 0.3679,
      "step": 44
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 6.751816749572754,
      "learning_rate": 2.25e-06,
      "loss": 0.2708,
      "step": 45
    },
    {
      "epoch": 1.0952380952380953,
      "grad_norm": 5.267884731292725,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.2143,
      "step": 46
    },
    {
      "epoch": 1.119047619047619,
      "grad_norm": 3.251101016998291,
      "learning_rate": 2.35e-06,
      "loss": 0.1759,
      "step": 47
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 2.585360050201416,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.1855,
      "step": 48
    },
    {
      "epoch": 1.1666666666666667,
      "grad_norm": 2.0107483863830566,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.1251,
      "step": 49
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 1.941689133644104,
      "learning_rate": 2.5e-06,
      "loss": 0.1441,
      "step": 50
    },
    {
      "epoch": 1.2142857142857142,
      "grad_norm": 1.4036344289779663,
      "learning_rate": 2.55e-06,
      "loss": 0.1097,
      "step": 51
    },
    {
      "epoch": 1.2380952380952381,
      "grad_norm": 1.2856179475784302,
      "learning_rate": 2.6e-06,
      "loss": 0.1303,
      "step": 52
    },
    {
      "epoch": 1.2619047619047619,
      "grad_norm": 1.3184994459152222,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.1021,
      "step": 53
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 1.144294261932373,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.115,
      "step": 54
    },
    {
      "epoch": 1.3095238095238095,
      "grad_norm": 0.9276831150054932,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.1028,
      "step": 55
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.9152742028236389,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0979,
      "step": 56
    },
    {
      "epoch": 1.3571428571428572,
      "grad_norm": 0.8525308966636658,
      "learning_rate": 2.85e-06,
      "loss": 0.0908,
      "step": 57
    },
    {
      "epoch": 1.380952380952381,
      "grad_norm": 0.9806348085403442,
      "learning_rate": 2.9e-06,
      "loss": 0.0817,
      "step": 58
    },
    {
      "epoch": 1.4047619047619047,
      "grad_norm": 0.606792151927948,
      "learning_rate": 2.95e-06,
      "loss": 0.0904,
      "step": 59
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.7274054884910583,
      "learning_rate": 3e-06,
      "loss": 0.0811,
      "step": 60
    },
    {
      "epoch": 1.4523809523809523,
      "grad_norm": 1.0523946285247803,
      "learning_rate": 3.05e-06,
      "loss": 0.0881,
      "step": 61
    },
    {
      "epoch": 1.4761904761904763,
      "grad_norm": 0.5840473175048828,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0848,
      "step": 62
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.7410831451416016,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0788,
      "step": 63
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.828996479511261,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0866,
      "step": 64
    },
    {
      "epoch": 1.5476190476190477,
      "grad_norm": 0.7505109310150146,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.076,
      "step": 65
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.9672189354896545,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0765,
      "step": 66
    },
    {
      "epoch": 1.5952380952380953,
      "grad_norm": 0.5925746560096741,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0706,
      "step": 67
    },
    {
      "epoch": 1.619047619047619,
      "grad_norm": 0.6671133637428284,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0713,
      "step": 68
    },
    {
      "epoch": 1.6428571428571428,
      "grad_norm": 0.5542609095573425,
      "learning_rate": 3.45e-06,
      "loss": 0.0715,
      "step": 69
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.541200578212738,
      "learning_rate": 3.5e-06,
      "loss": 0.0701,
      "step": 70
    },
    {
      "epoch": 1.6904761904761905,
      "grad_norm": 0.4222320020198822,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0669,
      "step": 71
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.717410147190094,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0782,
      "step": 72
    },
    {
      "epoch": 1.7380952380952381,
      "grad_norm": 0.6776471734046936,
      "learning_rate": 3.65e-06,
      "loss": 0.0695,
      "step": 73
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 0.5480474829673767,
      "learning_rate": 3.7e-06,
      "loss": 0.0662,
      "step": 74
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.4779343605041504,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0743,
      "step": 75
    },
    {
      "epoch": 1.8095238095238095,
      "grad_norm": 0.43138471245765686,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0666,
      "step": 76
    },
    {
      "epoch": 1.8333333333333335,
      "grad_norm": 0.6058762669563293,
      "learning_rate": 3.85e-06,
      "loss": 0.0696,
      "step": 77
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 1.3352755308151245,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0891,
      "step": 78
    },
    {
      "epoch": 1.880952380952381,
      "grad_norm": 0.5319089293479919,
      "learning_rate": 3.95e-06,
      "loss": 0.0617,
      "step": 79
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.5629184246063232,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0622,
      "step": 80
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 0.37953704595565796,
      "learning_rate": 4.05e-06,
      "loss": 0.0676,
      "step": 81
    },
    {
      "epoch": 1.9523809523809523,
      "grad_norm": 0.37576770782470703,
      "learning_rate": 4.1e-06,
      "loss": 0.0719,
      "step": 82
    },
    {
      "epoch": 1.9761904761904763,
      "grad_norm": 0.4720636010169983,
      "learning_rate": 4.15e-06,
      "loss": 0.0662,
      "step": 83
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4793304204940796,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0664,
      "step": 84
    },
    {
      "epoch": 2.0238095238095237,
      "grad_norm": 0.44540268182754517,
      "learning_rate": 4.25e-06,
      "loss": 0.0627,
      "step": 85
    },
    {
      "epoch": 2.0476190476190474,
      "grad_norm": 0.40029338002204895,
      "learning_rate": 4.3e-06,
      "loss": 0.0615,
      "step": 86
    },
    {
      "epoch": 2.0714285714285716,
      "grad_norm": 0.44975095987319946,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0556,
      "step": 87
    },
    {
      "epoch": 2.0952380952380953,
      "grad_norm": 0.49270153045654297,
      "learning_rate": 4.4e-06,
      "loss": 0.0509,
      "step": 88
    },
    {
      "epoch": 2.119047619047619,
      "grad_norm": 0.5457497239112854,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0654,
      "step": 89
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.5662055611610413,
      "learning_rate": 4.5e-06,
      "loss": 0.0584,
      "step": 90
    },
    {
      "epoch": 2.1666666666666665,
      "grad_norm": 0.4589175581932068,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.05,
      "step": 91
    },
    {
      "epoch": 2.1904761904761907,
      "grad_norm": 0.4754093289375305,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0409,
      "step": 92
    },
    {
      "epoch": 2.2142857142857144,
      "grad_norm": 0.5139522552490234,
      "learning_rate": 4.65e-06,
      "loss": 0.0548,
      "step": 93
    },
    {
      "epoch": 2.238095238095238,
      "grad_norm": 0.6404117345809937,
      "learning_rate": 4.7e-06,
      "loss": 0.0575,
      "step": 94
    },
    {
      "epoch": 2.261904761904762,
      "grad_norm": 0.4840497672557831,
      "learning_rate": 4.75e-06,
      "loss": 0.0545,
      "step": 95
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.468397319316864,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0532,
      "step": 96
    },
    {
      "epoch": 2.3095238095238093,
      "grad_norm": 0.41116851568222046,
      "learning_rate": 4.85e-06,
      "loss": 0.0508,
      "step": 97
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.4745863080024719,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0571,
      "step": 98
    },
    {
      "epoch": 2.357142857142857,
      "grad_norm": 0.5628818869590759,
      "learning_rate": 4.95e-06,
      "loss": 0.0527,
      "step": 99
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.5258535146713257,
      "learning_rate": 5e-06,
      "loss": 0.0602,
      "step": 100
    },
    {
      "epoch": 2.4047619047619047,
      "grad_norm": 0.4047287702560425,
      "learning_rate": 4.999466041969828e-06,
      "loss": 0.0511,
      "step": 101
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.493963360786438,
      "learning_rate": 4.997864395968252e-06,
      "loss": 0.0496,
      "step": 102
    },
    {
      "epoch": 2.4523809523809526,
      "grad_norm": 0.4637805223464966,
      "learning_rate": 4.9951957461646705e-06,
      "loss": 0.047,
      "step": 103
    },
    {
      "epoch": 2.4761904761904763,
      "grad_norm": 0.4990857243537903,
      "learning_rate": 4.991461232516675e-06,
      "loss": 0.0566,
      "step": 104
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.41547098755836487,
      "learning_rate": 4.986662450283107e-06,
      "loss": 0.0496,
      "step": 105
    },
    {
      "epoch": 2.5238095238095237,
      "grad_norm": 0.5748231410980225,
      "learning_rate": 4.9808014493426124e-06,
      "loss": 0.0362,
      "step": 106
    },
    {
      "epoch": 2.5476190476190474,
      "grad_norm": 0.481645792722702,
      "learning_rate": 4.973880733318007e-06,
      "loss": 0.0418,
      "step": 107
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.49727940559387207,
      "learning_rate": 4.965903258506806e-06,
      "loss": 0.0436,
      "step": 108
    },
    {
      "epoch": 2.5952380952380953,
      "grad_norm": 0.6753897070884705,
      "learning_rate": 4.956872432618399e-06,
      "loss": 0.0486,
      "step": 109
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.6681185960769653,
      "learning_rate": 4.9467921133183864e-06,
      "loss": 0.0426,
      "step": 110
    },
    {
      "epoch": 2.642857142857143,
      "grad_norm": 0.6316296458244324,
      "learning_rate": 4.935666606580719e-06,
      "loss": 0.0377,
      "step": 111
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.43960002064704895,
      "learning_rate": 4.923500664848327e-06,
      "loss": 0.0355,
      "step": 112
    },
    {
      "epoch": 2.6904761904761907,
      "grad_norm": 0.437860906124115,
      "learning_rate": 4.910299485003034e-06,
      "loss": 0.0369,
      "step": 113
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 0.5523399710655212,
      "learning_rate": 4.896068706145632e-06,
      "loss": 0.0488,
      "step": 114
    },
    {
      "epoch": 2.738095238095238,
      "grad_norm": 0.6876885890960693,
      "learning_rate": 4.880814407187037e-06,
      "loss": 0.0492,
      "step": 115
    },
    {
      "epoch": 2.761904761904762,
      "grad_norm": 0.48951950669288635,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.0464,
      "step": 116
    },
    {
      "epoch": 2.7857142857142856,
      "grad_norm": 0.6003612875938416,
      "learning_rate": 4.8472617478935744e-06,
      "loss": 0.0376,
      "step": 117
    },
    {
      "epoch": 2.8095238095238093,
      "grad_norm": 0.5665768384933472,
      "learning_rate": 4.828977720128198e-06,
      "loss": 0.0364,
      "step": 118
    },
    {
      "epoch": 2.8333333333333335,
      "grad_norm": 0.6751538515090942,
      "learning_rate": 4.809698831278217e-06,
      "loss": 0.0415,
      "step": 119
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6295377612113953,
      "learning_rate": 4.789433316637644e-06,
      "loss": 0.0364,
      "step": 120
    },
    {
      "epoch": 2.880952380952381,
      "grad_norm": 0.4297006130218506,
      "learning_rate": 4.7681898329539004e-06,
      "loss": 0.0282,
      "step": 121
    },
    {
      "epoch": 2.9047619047619047,
      "grad_norm": 0.48039913177490234,
      "learning_rate": 4.745977454729947e-06,
      "loss": 0.0319,
      "step": 122
    },
    {
      "epoch": 2.928571428571429,
      "grad_norm": 0.5199178457260132,
      "learning_rate": 4.722805670347963e-06,
      "loss": 0.0299,
      "step": 123
    },
    {
      "epoch": 2.9523809523809526,
      "grad_norm": 0.5075429677963257,
      "learning_rate": 4.698684378016223e-06,
      "loss": 0.0265,
      "step": 124
    },
    {
      "epoch": 2.9761904761904763,
      "grad_norm": 0.5364718437194824,
      "learning_rate": 4.673623881540917e-06,
      "loss": 0.0351,
      "step": 125
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5693520903587341,
      "learning_rate": 4.647634885924713e-06,
      "loss": 0.0286,
      "step": 126
    },
    {
      "epoch": 3.0238095238095237,
      "grad_norm": 0.502586305141449,
      "learning_rate": 4.620728492793934e-06,
      "loss": 0.0222,
      "step": 127
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.42680972814559937,
      "learning_rate": 4.592916195656322e-06,
      "loss": 0.0183,
      "step": 128
    },
    {
      "epoch": 3.0714285714285716,
      "grad_norm": 0.6365283131599426,
      "learning_rate": 4.56420987499139e-06,
      "loss": 0.025,
      "step": 129
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 0.4085887670516968,
      "learning_rate": 4.534621793175488e-06,
      "loss": 0.0166,
      "step": 130
    },
    {
      "epoch": 3.119047619047619,
      "grad_norm": 0.4094793498516083,
      "learning_rate": 4.504164589243721e-06,
      "loss": 0.0158,
      "step": 131
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.6991594433784485,
      "learning_rate": 4.472851273490985e-06,
      "loss": 0.0253,
      "step": 132
    },
    {
      "epoch": 3.1666666666666665,
      "grad_norm": 0.4174569547176361,
      "learning_rate": 4.440695221914394e-06,
      "loss": 0.0165,
      "step": 133
    },
    {
      "epoch": 3.1904761904761907,
      "grad_norm": 0.4691311717033386,
      "learning_rate": 4.407710170499517e-06,
      "loss": 0.0187,
      "step": 134
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 0.6574212312698364,
      "learning_rate": 4.373910209352816e-06,
      "loss": 0.018,
      "step": 135
    },
    {
      "epoch": 3.238095238095238,
      "grad_norm": 0.40424057841300964,
      "learning_rate": 4.33930977668283e-06,
      "loss": 0.0115,
      "step": 136
    },
    {
      "epoch": 3.261904761904762,
      "grad_norm": 0.5574136972427368,
      "learning_rate": 4.303923652632656e-06,
      "loss": 0.0143,
      "step": 137
    },
    {
      "epoch": 3.2857142857142856,
      "grad_norm": 0.7671014666557312,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.0123,
      "step": 138
    },
    {
      "epoch": 3.3095238095238093,
      "grad_norm": 0.5041195750236511,
      "learning_rate": 4.2308551226120745e-06,
      "loss": 0.0105,
      "step": 139
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.7552913427352905,
      "learning_rate": 4.1932039290643534e-06,
      "loss": 0.0132,
      "step": 140
    },
    {
      "epoch": 3.357142857142857,
      "grad_norm": 0.4549916982650757,
      "learning_rate": 4.154829455648916e-06,
      "loss": 0.0073,
      "step": 141
    },
    {
      "epoch": 3.380952380952381,
      "grad_norm": 0.5453060269355774,
      "learning_rate": 4.115748094652352e-06,
      "loss": 0.0065,
      "step": 142
    },
    {
      "epoch": 3.4047619047619047,
      "grad_norm": 0.6225312948226929,
      "learning_rate": 4.075976540319888e-06,
      "loss": 0.0087,
      "step": 143
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.629115641117096,
      "learning_rate": 4.0355317817241705e-06,
      "loss": 0.0118,
      "step": 144
    },
    {
      "epoch": 3.4523809523809526,
      "grad_norm": 0.537651538848877,
      "learning_rate": 3.994431095508102e-06,
      "loss": 0.0066,
      "step": 145
    },
    {
      "epoch": 3.4761904761904763,
      "grad_norm": 0.9304003715515137,
      "learning_rate": 3.9526920385048465e-06,
      "loss": 0.0082,
      "step": 146
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.557716965675354,
      "learning_rate": 3.9103324402381285e-06,
      "loss": 0.0063,
      "step": 147
    },
    {
      "epoch": 3.5238095238095237,
      "grad_norm": 0.3688383102416992,
      "learning_rate": 3.8673703953060685e-06,
      "loss": 0.0049,
      "step": 148
    },
    {
      "epoch": 3.5476190476190474,
      "grad_norm": 0.44352808594703674,
      "learning_rate": 3.8238242556517725e-06,
      "loss": 0.0043,
      "step": 149
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.5826281905174255,
      "learning_rate": 3.779712622724003e-06,
      "loss": 0.0063,
      "step": 150
    },
    {
      "epoch": 3.5952380952380953,
      "grad_norm": 0.600183367729187,
      "learning_rate": 3.7350543395312604e-06,
      "loss": 0.0057,
      "step": 151
    },
    {
      "epoch": 3.619047619047619,
      "grad_norm": 0.6133769750595093,
      "learning_rate": 3.6898684825926845e-06,
      "loss": 0.007,
      "step": 152
    },
    {
      "epoch": 3.642857142857143,
      "grad_norm": 0.5637931227684021,
      "learning_rate": 3.6441743537892045e-06,
      "loss": 0.0059,
      "step": 153
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.27519500255584717,
      "learning_rate": 3.5979914721184263e-06,
      "loss": 0.002,
      "step": 154
    },
    {
      "epoch": 3.6904761904761907,
      "grad_norm": 0.3693721294403076,
      "learning_rate": 3.551339565356769e-06,
      "loss": 0.0042,
      "step": 155
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.733023464679718,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.0031,
      "step": 156
    },
    {
      "epoch": 3.738095238095238,
      "grad_norm": 0.4825158417224884,
      "learning_rate": 3.4567085809127247e-06,
      "loss": 0.0034,
      "step": 157
    },
    {
      "epoch": 3.761904761904762,
      "grad_norm": 0.745896577835083,
      "learning_rate": 3.4087699264095746e-06,
      "loss": 0.006,
      "step": 158
    },
    {
      "epoch": 3.7857142857142856,
      "grad_norm": 0.606246829032898,
      "learning_rate": 3.360443075906597e-06,
      "loss": 0.0028,
      "step": 159
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.35037049651145935,
      "learning_rate": 3.3117486730117092e-06,
      "loss": 0.0022,
      "step": 160
    },
    {
      "epoch": 3.8333333333333335,
      "grad_norm": 0.3873816132545471,
      "learning_rate": 3.2627075183388725e-06,
      "loss": 0.0024,
      "step": 161
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 0.2003440409898758,
      "learning_rate": 3.2133405606227636e-06,
      "loss": 0.0014,
      "step": 162
    },
    {
      "epoch": 3.880952380952381,
      "grad_norm": 1.2881035804748535,
      "learning_rate": 3.163668887770181e-06,
      "loss": 0.0052,
      "step": 163
    },
    {
      "epoch": 3.9047619047619047,
      "grad_norm": 0.16532732546329498,
      "learning_rate": 3.1137137178519983e-06,
      "loss": 0.0006,
      "step": 164
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 0.42558538913726807,
      "learning_rate": 3.063496390039516e-06,
      "loss": 0.0018,
      "step": 165
    },
    {
      "epoch": 3.9523809523809526,
      "grad_norm": 0.5848471522331238,
      "learning_rate": 3.013038355489086e-06,
      "loss": 0.002,
      "step": 166
    },
    {
      "epoch": 3.9761904761904763,
      "grad_norm": 0.18496190011501312,
      "learning_rate": 2.9623611681788967e-06,
      "loss": 0.0012,
      "step": 167
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9864285588264465,
      "learning_rate": 2.911486475701835e-06,
      "loss": 0.0021,
      "step": 168
    },
    {
      "epoch": 4.023809523809524,
      "grad_norm": 0.28038740158081055,
      "learning_rate": 2.860436010018367e-06,
      "loss": 0.0016,
      "step": 169
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.174965500831604,
      "learning_rate": 2.80923157817337e-06,
      "loss": 0.0006,
      "step": 170
    },
    {
      "epoch": 4.071428571428571,
      "grad_norm": 0.111990787088871,
      "learning_rate": 2.7578950529808927e-06,
      "loss": 0.0003,
      "step": 171
    },
    {
      "epoch": 4.095238095238095,
      "grad_norm": 0.0902976393699646,
      "learning_rate": 2.7064483636808314e-06,
      "loss": 0.0004,
      "step": 172
    },
    {
      "epoch": 4.119047619047619,
      "grad_norm": 0.23398929834365845,
      "learning_rate": 2.654913486571487e-06,
      "loss": 0.0007,
      "step": 173
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 0.3977113664150238,
      "learning_rate": 2.603312435622033e-06,
      "loss": 0.0012,
      "step": 174
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.18411217629909515,
      "learning_rate": 2.5516672530688864e-06,
      "loss": 0.0006,
      "step": 175
    },
    {
      "epoch": 4.190476190476191,
      "grad_norm": 0.3521343171596527,
      "learning_rate": 2.5e-06,
      "loss": 0.0017,
      "step": 176
    },
    {
      "epoch": 4.214285714285714,
      "grad_norm": 0.1081409603357315,
      "learning_rate": 2.448332746931115e-06,
      "loss": 0.0004,
      "step": 177
    },
    {
      "epoch": 4.238095238095238,
      "grad_norm": 0.10767076164484024,
      "learning_rate": 2.396687564377967e-06,
      "loss": 0.0003,
      "step": 178
    },
    {
      "epoch": 4.261904761904762,
      "grad_norm": 0.0657818615436554,
      "learning_rate": 2.345086513428514e-06,
      "loss": 0.0003,
      "step": 179
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.03795298561453819,
      "learning_rate": 2.2935516363191695e-06,
      "loss": 0.0002,
      "step": 180
    },
    {
      "epoch": 4.309523809523809,
      "grad_norm": 0.09580468386411667,
      "learning_rate": 2.2421049470191077e-06,
      "loss": 0.0002,
      "step": 181
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.13761642575263977,
      "learning_rate": 2.190768421826631e-06,
      "loss": 0.0004,
      "step": 182
    },
    {
      "epoch": 4.357142857142857,
      "grad_norm": 0.10877614468336105,
      "learning_rate": 2.139563989981633e-06,
      "loss": 0.0003,
      "step": 183
    },
    {
      "epoch": 4.380952380952381,
      "grad_norm": 0.14662612974643707,
      "learning_rate": 2.088513524298165e-06,
      "loss": 0.0003,
      "step": 184
    },
    {
      "epoch": 4.404761904761905,
      "grad_norm": 0.29953065514564514,
      "learning_rate": 2.037638831821104e-06,
      "loss": 0.0004,
      "step": 185
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.20808857679367065,
      "learning_rate": 1.9869616445109146e-06,
      "loss": 0.0003,
      "step": 186
    },
    {
      "epoch": 4.4523809523809526,
      "grad_norm": 0.03407036140561104,
      "learning_rate": 1.9365036099604853e-06,
      "loss": 0.0002,
      "step": 187
    },
    {
      "epoch": 4.476190476190476,
      "grad_norm": 0.3132052421569824,
      "learning_rate": 1.8862862821480023e-06,
      "loss": 0.0006,
      "step": 188
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.2567755877971649,
      "learning_rate": 1.83633111222982e-06,
      "loss": 0.0007,
      "step": 189
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.11175425350666046,
      "learning_rate": 1.7866594393772375e-06,
      "loss": 0.0003,
      "step": 190
    },
    {
      "epoch": 4.5476190476190474,
      "grad_norm": 0.594926118850708,
      "learning_rate": 1.7372924816611283e-06,
      "loss": 0.001,
      "step": 191
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.03871627524495125,
      "learning_rate": 1.6882513269882916e-06,
      "loss": 0.0002,
      "step": 192
    },
    {
      "epoch": 4.595238095238095,
      "grad_norm": 0.06972040235996246,
      "learning_rate": 1.6395569240934042e-06,
      "loss": 0.0002,
      "step": 193
    },
    {
      "epoch": 4.619047619047619,
      "grad_norm": 0.057279668748378754,
      "learning_rate": 1.5912300735904252e-06,
      "loss": 0.0002,
      "step": 194
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 0.09757455438375473,
      "learning_rate": 1.5432914190872757e-06,
      "loss": 0.0003,
      "step": 195
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.1020384356379509,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.0003,
      "step": 196
    },
    {
      "epoch": 4.690476190476191,
      "grad_norm": 0.08006434142589569,
      "learning_rate": 1.4486604346432311e-06,
      "loss": 0.0003,
      "step": 197
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 0.0277456846088171,
      "learning_rate": 1.4020085278815745e-06,
      "loss": 0.0001,
      "step": 198
    },
    {
      "epoch": 4.738095238095238,
      "grad_norm": 0.018923059105873108,
      "learning_rate": 1.3558256462107965e-06,
      "loss": 0.0002,
      "step": 199
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.01575353741645813,
      "learning_rate": 1.3101315174073162e-06,
      "loss": 0.0001,
      "step": 200
    },
    {
      "epoch": 4.785714285714286,
      "grad_norm": 0.014921327121555805,
      "learning_rate": 1.2649456604687404e-06,
      "loss": 0.0002,
      "step": 201
    },
    {
      "epoch": 4.809523809523809,
      "grad_norm": 0.03840240091085434,
      "learning_rate": 1.2202873772759983e-06,
      "loss": 0.0002,
      "step": 202
    },
    {
      "epoch": 4.833333333333333,
      "grad_norm": 0.04131830111145973,
      "learning_rate": 1.1761757443482285e-06,
      "loss": 0.0002,
      "step": 203
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.02138919197022915,
      "learning_rate": 1.1326296046939334e-06,
      "loss": 0.0002,
      "step": 204
    },
    {
      "epoch": 4.880952380952381,
      "grad_norm": 0.016792714595794678,
      "learning_rate": 1.0896675597618725e-06,
      "loss": 0.0001,
      "step": 205
    },
    {
      "epoch": 4.904761904761905,
      "grad_norm": 0.017704278230667114,
      "learning_rate": 1.0473079614951546e-06,
      "loss": 0.0002,
      "step": 206
    },
    {
      "epoch": 4.928571428571429,
      "grad_norm": 0.020070459693670273,
      "learning_rate": 1.0055689044918979e-06,
      "loss": 0.0001,
      "step": 207
    },
    {
      "epoch": 4.9523809523809526,
      "grad_norm": 0.035451311618089676,
      "learning_rate": 9.644682182758305e-07,
      "loss": 0.0002,
      "step": 208
    },
    {
      "epoch": 4.976190476190476,
      "grad_norm": 0.02921333909034729,
      "learning_rate": 9.240234596801125e-07,
      "loss": 0.0002,
      "step": 209
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.006285260431468487,
      "learning_rate": 8.842519053476476e-07,
      "loss": 0.0001,
      "step": 210
    },
    {
      "epoch": 5.023809523809524,
      "grad_norm": 0.01432103756815195,
      "learning_rate": 8.451705443510838e-07,
      "loss": 0.0001,
      "step": 211
    },
    {
      "epoch": 5.0476190476190474,
      "grad_norm": 0.11769424378871918,
      "learning_rate": 8.067960709356479e-07,
      "loss": 0.0002,
      "step": 212
    },
    {
      "epoch": 5.071428571428571,
      "grad_norm": 0.01345640979707241,
      "learning_rate": 7.691448773879257e-07,
      "loss": 0.0002,
      "step": 213
    },
    {
      "epoch": 5.095238095238095,
      "grad_norm": 0.023673633113503456,
      "learning_rate": 7.322330470336314e-07,
      "loss": 0.0001,
      "step": 214
    },
    {
      "epoch": 5.119047619047619,
      "grad_norm": 0.014196817763149738,
      "learning_rate": 6.960763473673451e-07,
      "loss": 0.0001,
      "step": 215
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.033266760408878326,
      "learning_rate": 6.60690223317171e-07,
      "loss": 0.0002,
      "step": 216
    },
    {
      "epoch": 5.166666666666667,
      "grad_norm": 0.007004464510828257,
      "learning_rate": 6.260897906471852e-07,
      "loss": 0.0001,
      "step": 217
    },
    {
      "epoch": 5.190476190476191,
      "grad_norm": 0.008027388714253902,
      "learning_rate": 5.922898295004842e-07,
      "loss": 0.0001,
      "step": 218
    },
    {
      "epoch": 5.214285714285714,
      "grad_norm": 0.030500022694468498,
      "learning_rate": 5.593047780856065e-07,
      "loss": 0.0002,
      "step": 219
    },
    {
      "epoch": 5.238095238095238,
      "grad_norm": 0.00899834930896759,
      "learning_rate": 5.271487265090163e-07,
      "loss": 0.0001,
      "step": 220
    },
    {
      "epoch": 5.261904761904762,
      "grad_norm": 0.02572966367006302,
      "learning_rate": 4.958354107562793e-07,
      "loss": 0.0001,
      "step": 221
    },
    {
      "epoch": 5.285714285714286,
      "grad_norm": 0.023606926202774048,
      "learning_rate": 4.6537820682451273e-07,
      "loss": 0.0002,
      "step": 222
    },
    {
      "epoch": 5.309523809523809,
      "grad_norm": 0.01611892133951187,
      "learning_rate": 4.3579012500861077e-07,
      "loss": 0.0001,
      "step": 223
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.011451052501797676,
      "learning_rate": 4.070838043436787e-07,
      "loss": 0.0001,
      "step": 224
    },
    {
      "epoch": 5.357142857142857,
      "grad_norm": 0.014691251330077648,
      "learning_rate": 3.7927150720606596e-07,
      "loss": 0.0002,
      "step": 225
    },
    {
      "epoch": 5.380952380952381,
      "grad_norm": 0.010912562720477581,
      "learning_rate": 3.5236511407528676e-07,
      "loss": 0.0001,
      "step": 226
    },
    {
      "epoch": 5.404761904761905,
      "grad_norm": 0.031578365713357925,
      "learning_rate": 3.2637611845908274e-07,
      "loss": 0.0001,
      "step": 227
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.012554287910461426,
      "learning_rate": 3.0131562198377763e-07,
      "loss": 0.0001,
      "step": 228
    },
    {
      "epoch": 5.4523809523809526,
      "grad_norm": 0.012699681334197521,
      "learning_rate": 2.7719432965203685e-07,
      "loss": 0.0002,
      "step": 229
    },
    {
      "epoch": 5.476190476190476,
      "grad_norm": 0.014920556917786598,
      "learning_rate": 2.5402254527005286e-07,
      "loss": 0.0001,
      "step": 230
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.01147391740232706,
      "learning_rate": 2.3181016704610038e-07,
      "loss": 0.0001,
      "step": 231
    },
    {
      "epoch": 5.523809523809524,
      "grad_norm": 0.01795969158411026,
      "learning_rate": 2.1056668336235624e-07,
      "loss": 0.0001,
      "step": 232
    },
    {
      "epoch": 5.5476190476190474,
      "grad_norm": 0.02227536402642727,
      "learning_rate": 1.9030116872178317e-07,
      "loss": 0.0001,
      "step": 233
    },
    {
      "epoch": 5.571428571428571,
      "grad_norm": 0.010887685231864452,
      "learning_rate": 1.710222798718028e-07,
      "loss": 0.0001,
      "step": 234
    },
    {
      "epoch": 5.595238095238095,
      "grad_norm": 0.011158541776239872,
      "learning_rate": 1.5273825210642608e-07,
      "loss": 0.0001,
      "step": 235
    },
    {
      "epoch": 5.619047619047619,
      "grad_norm": 0.008868343196809292,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.0001,
      "step": 236
    },
    {
      "epoch": 5.642857142857143,
      "grad_norm": 0.013471647165715694,
      "learning_rate": 1.1918559281296399e-07,
      "loss": 0.0001,
      "step": 237
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.009484678506851196,
      "learning_rate": 1.0393129385436824e-07,
      "loss": 0.0001,
      "step": 238
    },
    {
      "epoch": 5.690476190476191,
      "grad_norm": 0.022991865873336792,
      "learning_rate": 8.970051499696597e-08,
      "loss": 0.0001,
      "step": 239
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.00791236013174057,
      "learning_rate": 7.649933515167407e-08,
      "loss": 0.0001,
      "step": 240
    },
    {
      "epoch": 5.738095238095238,
      "grad_norm": 0.013699101284146309,
      "learning_rate": 6.433339341928136e-08,
      "loss": 0.0001,
      "step": 241
    },
    {
      "epoch": 5.761904761904762,
      "grad_norm": 0.01205279678106308,
      "learning_rate": 5.3207886681613804e-08,
      "loss": 0.0001,
      "step": 242
    },
    {
      "epoch": 5.785714285714286,
      "grad_norm": 0.007536980323493481,
      "learning_rate": 4.312756738160146e-08,
      "loss": 0.0001,
      "step": 243
    },
    {
      "epoch": 5.809523809523809,
      "grad_norm": 0.006828755605965853,
      "learning_rate": 3.4096741493194196e-08,
      "loss": 0.0001,
      "step": 244
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.010486593469977379,
      "learning_rate": 2.611926668199316e-08,
      "loss": 0.0001,
      "step": 245
    },
    {
      "epoch": 5.857142857142857,
      "grad_norm": 0.012378604151308537,
      "learning_rate": 1.919855065738746e-08,
      "loss": 0.0001,
      "step": 246
    },
    {
      "epoch": 5.880952380952381,
      "grad_norm": 0.011314993724226952,
      "learning_rate": 1.3337549716893606e-08,
      "loss": 0.0001,
      "step": 247
    },
    {
      "epoch": 5.904761904761905,
      "grad_norm": 0.00836703460663557,
      "learning_rate": 8.538767483325384e-09,
      "loss": 0.0001,
      "step": 248
    },
    {
      "epoch": 5.928571428571429,
      "grad_norm": 0.00872640311717987,
      "learning_rate": 4.8042538353296195e-09,
      "loss": 0.0001,
      "step": 249
    },
    {
      "epoch": 5.9523809523809526,
      "grad_norm": 0.014776388183236122,
      "learning_rate": 2.1356040317474512e-09,
      "loss": 0.0001,
      "step": 250
    },
    {
      "epoch": 5.976190476190476,
      "grad_norm": 0.01690945215523243,
      "learning_rate": 5.339580301724767e-10,
      "loss": 0.0001,
      "step": 251
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02883993461728096,
      "learning_rate": 0.0,
      "loss": 0.0001,
      "step": 252
    }
  ],
  "logging_steps": 1,
  "max_steps": 252,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 42,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.645160370101289e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}