{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9986013986013986,
  "eval_steps": 500,
  "global_step": 238,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004195804195804196,
      "grad_norm": 717.93115234375,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 2.7062,
      "step": 1
    },
    {
      "epoch": 0.008391608391608392,
      "grad_norm": 700.4446411132812,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 2.7012,
      "step": 2
    },
    {
      "epoch": 0.012587412587412588,
      "grad_norm": 1356.7574462890625,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 2.7212,
      "step": 3
    },
    {
      "epoch": 0.016783216783216783,
      "grad_norm": 899.5720825195312,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.7709,
      "step": 4
    },
    {
      "epoch": 0.02097902097902098,
      "grad_norm": 875.39892578125,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 2.7834,
      "step": 5
    },
    {
      "epoch": 0.025174825174825177,
      "grad_norm": 925.267822265625,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 2.7798,
      "step": 6
    },
    {
      "epoch": 0.02937062937062937,
      "grad_norm": 71.72831726074219,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 2.774,
      "step": 7
    },
    {
      "epoch": 0.033566433566433566,
      "grad_norm": 569.9453735351562,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.7603,
      "step": 8
    },
    {
      "epoch": 0.03776223776223776,
      "grad_norm": 384.319580078125,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 2.7579,
      "step": 9
    },
    {
      "epoch": 0.04195804195804196,
      "grad_norm": 599.7899169921875,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.709,
      "step": 10
    },
    {
      "epoch": 0.046153846153846156,
      "grad_norm": 566.8663940429688,
      "learning_rate": 5.5e-07,
      "loss": 2.6726,
      "step": 11
    },
    {
      "epoch": 0.05034965034965035,
      "grad_norm": 917.609619140625,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.6404,
      "step": 12
    },
    {
      "epoch": 0.05454545454545454,
      "grad_norm": 1377.7421875,
      "learning_rate": 6.5e-07,
      "loss": 2.6995,
      "step": 13
    },
    {
      "epoch": 0.05874125874125874,
      "grad_norm": 1760.7694091796875,
      "learning_rate": 7.000000000000001e-07,
      "loss": 2.6683,
      "step": 14
    },
    {
      "epoch": 0.06293706293706294,
      "grad_norm": 1408.8568115234375,
      "learning_rate": 7.5e-07,
      "loss": 2.4465,
      "step": 15
    },
    {
      "epoch": 0.06713286713286713,
      "grad_norm": 140.02610778808594,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.4905,
      "step": 16
    },
    {
      "epoch": 0.07132867132867132,
      "grad_norm": 243.03379821777344,
      "learning_rate": 8.500000000000001e-07,
      "loss": 2.5882,
      "step": 17
    },
    {
      "epoch": 0.07552447552447553,
      "grad_norm": 1498.6507568359375,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.4565,
      "step": 18
    },
    {
      "epoch": 0.07972027972027972,
      "grad_norm": 1193.4425048828125,
      "learning_rate": 9.500000000000001e-07,
      "loss": 2.4394,
      "step": 19
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 1025.7431640625,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.3618,
      "step": 20
    },
    {
      "epoch": 0.08811188811188811,
      "grad_norm": 194.40928649902344,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 2.2295,
      "step": 21
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 509.2408752441406,
      "learning_rate": 1.1e-06,
      "loss": 2.1498,
      "step": 22
    },
    {
      "epoch": 0.0965034965034965,
      "grad_norm": 177.00662231445312,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 2.136,
      "step": 23
    },
    {
      "epoch": 0.1006993006993007,
      "grad_norm": 266.7733459472656,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.0781,
      "step": 24
    },
    {
      "epoch": 0.1048951048951049,
      "grad_norm": 515.4574584960938,
      "learning_rate": 1.25e-06,
      "loss": 2.1255,
      "step": 25
    },
    {
      "epoch": 0.10909090909090909,
      "grad_norm": 242.0486297607422,
      "learning_rate": 1.3e-06,
      "loss": 1.9491,
      "step": 26
    },
    {
      "epoch": 0.11328671328671329,
      "grad_norm": 226.7819366455078,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 1.7425,
      "step": 27
    },
    {
      "epoch": 0.11748251748251748,
      "grad_norm": 657.841064453125,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 1.7396,
      "step": 28
    },
    {
      "epoch": 0.12167832167832168,
      "grad_norm": 2119.27197265625,
      "learning_rate": 1.45e-06,
      "loss": 1.6205,
      "step": 29
    },
    {
      "epoch": 0.1258741258741259,
      "grad_norm": 2379.28857421875,
      "learning_rate": 1.5e-06,
      "loss": 1.5446,
      "step": 30
    },
    {
      "epoch": 0.13006993006993006,
      "grad_norm": 306.68927001953125,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 1.4194,
      "step": 31
    },
    {
      "epoch": 0.13426573426573427,
      "grad_norm": 653.746337890625,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.3444,
      "step": 32
    },
    {
      "epoch": 0.13846153846153847,
      "grad_norm": 886.5042114257812,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 1.2742,
      "step": 33
    },
    {
      "epoch": 0.14265734265734265,
      "grad_norm": 852.1470947265625,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.1457,
      "step": 34
    },
    {
      "epoch": 0.14685314685314685,
      "grad_norm": 1664.3746337890625,
      "learning_rate": 1.75e-06,
      "loss": 1.164,
      "step": 35
    },
    {
      "epoch": 0.15104895104895105,
      "grad_norm": 1804.9078369140625,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.1269,
      "step": 36
    },
    {
      "epoch": 0.15524475524475526,
      "grad_norm": 167.39125061035156,
      "learning_rate": 1.85e-06,
      "loss": 0.9994,
      "step": 37
    },
    {
      "epoch": 0.15944055944055943,
      "grad_norm": 397.78350830078125,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.8804,
      "step": 38
    },
    {
      "epoch": 0.16363636363636364,
      "grad_norm": 1123.1029052734375,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.8312,
      "step": 39
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 1532.86474609375,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7654,
      "step": 40
    },
    {
      "epoch": 0.17202797202797201,
      "grad_norm": 755.2657470703125,
      "learning_rate": 2.05e-06,
      "loss": 0.6198,
      "step": 41
    },
    {
      "epoch": 0.17622377622377622,
      "grad_norm": 1052.9649658203125,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.5466,
      "step": 42
    },
    {
      "epoch": 0.18041958041958042,
      "grad_norm": 219.14976501464844,
      "learning_rate": 2.15e-06,
      "loss": 0.4735,
      "step": 43
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 1271.767333984375,
      "learning_rate": 2.2e-06,
      "loss": 0.4669,
      "step": 44
    },
    {
      "epoch": 0.1888111888111888,
      "grad_norm": 1254.658935546875,
      "learning_rate": 2.25e-06,
      "loss": 0.3965,
      "step": 45
    },
    {
      "epoch": 0.193006993006993,
      "grad_norm": 927.441162109375,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.3704,
      "step": 46
    },
    {
      "epoch": 0.1972027972027972,
      "grad_norm": 424.604736328125,
      "learning_rate": 2.35e-06,
      "loss": 0.2998,
      "step": 47
    },
    {
      "epoch": 0.2013986013986014,
      "grad_norm": 562.2470703125,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.2485,
      "step": 48
    },
    {
      "epoch": 0.2055944055944056,
      "grad_norm": 619.4100341796875,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.2549,
      "step": 49
    },
    {
      "epoch": 0.2097902097902098,
      "grad_norm": 282.3805847167969,
      "learning_rate": 2.5e-06,
      "loss": 0.2081,
      "step": 50
    },
    {
      "epoch": 0.213986013986014,
      "grad_norm": 530.0814208984375,
      "learning_rate": 2.55e-06,
      "loss": 0.229,
      "step": 51
    },
    {
      "epoch": 0.21818181818181817,
      "grad_norm": 315.1905212402344,
      "learning_rate": 2.6e-06,
      "loss": 0.201,
      "step": 52
    },
    {
      "epoch": 0.22237762237762237,
      "grad_norm": 363.47918701171875,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.1797,
      "step": 53
    },
    {
      "epoch": 0.22657342657342658,
      "grad_norm": 485.228271484375,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.1743,
      "step": 54
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 332.8292236328125,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.1804,
      "step": 55
    },
    {
      "epoch": 0.23496503496503496,
      "grad_norm": 386.52447509765625,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.1663,
      "step": 56
    },
    {
      "epoch": 0.23916083916083916,
      "grad_norm": 368.46307373046875,
      "learning_rate": 2.85e-06,
      "loss": 0.1784,
      "step": 57
    },
    {
      "epoch": 0.24335664335664337,
      "grad_norm": 336.0594787597656,
      "learning_rate": 2.9e-06,
      "loss": 0.1637,
      "step": 58
    },
    {
      "epoch": 0.24755244755244754,
      "grad_norm": 34.440250396728516,
      "learning_rate": 2.95e-06,
      "loss": 0.1392,
      "step": 59
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 1019.0678100585938,
      "learning_rate": 3e-06,
      "loss": 0.1957,
      "step": 60
    },
    {
      "epoch": 0.25594405594405595,
      "grad_norm": 1919.9234619140625,
      "learning_rate": 3.05e-06,
      "loss": 0.2649,
      "step": 61
    },
    {
      "epoch": 0.2601398601398601,
      "grad_norm": 1502.3551025390625,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.255,
      "step": 62
    },
    {
      "epoch": 0.26433566433566436,
      "grad_norm": 880.5719604492188,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.1904,
      "step": 63
    },
    {
      "epoch": 0.26853146853146853,
      "grad_norm": 140.78419494628906,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.1275,
      "step": 64
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 280.4155578613281,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.1373,
      "step": 65
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 337.9070739746094,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.1642,
      "step": 66
    },
    {
      "epoch": 0.2811188811188811,
      "grad_norm": 331.3836669921875,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.152,
      "step": 67
    },
    {
      "epoch": 0.2853146853146853,
      "grad_norm": 184.1329803466797,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.1101,
      "step": 68
    },
    {
      "epoch": 0.2895104895104895,
      "grad_norm": 3.4701905250549316,
      "learning_rate": 3.45e-06,
      "loss": 0.1073,
      "step": 69
    },
    {
      "epoch": 0.2937062937062937,
      "grad_norm": 244.07217407226562,
      "learning_rate": 3.5e-06,
      "loss": 0.1375,
      "step": 70
    },
    {
      "epoch": 0.29790209790209793,
      "grad_norm": 304.42498779296875,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.1409,
      "step": 71
    },
    {
      "epoch": 0.3020979020979021,
      "grad_norm": 216.21617126464844,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.1348,
      "step": 72
    },
    {
      "epoch": 0.3062937062937063,
      "grad_norm": 105.47506713867188,
      "learning_rate": 3.65e-06,
      "loss": 0.113,
      "step": 73
    },
    {
      "epoch": 0.3104895104895105,
      "grad_norm": 75.4766845703125,
      "learning_rate": 3.7e-06,
      "loss": 0.1107,
      "step": 74
    },
    {
      "epoch": 0.3146853146853147,
      "grad_norm": 48.33282470703125,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.1177,
      "step": 75
    },
    {
      "epoch": 0.31888111888111886,
      "grad_norm": 9.441969871520996,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.1258,
      "step": 76
    },
    {
      "epoch": 0.3230769230769231,
      "grad_norm": 45.23430252075195,
      "learning_rate": 3.85e-06,
      "loss": 0.1016,
      "step": 77
    },
    {
      "epoch": 0.32727272727272727,
      "grad_norm": 53.305484771728516,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.1054,
      "step": 78
    },
    {
      "epoch": 0.33146853146853145,
      "grad_norm": 27.8891544342041,
      "learning_rate": 3.95e-06,
      "loss": 0.1066,
      "step": 79
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 18.898664474487305,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.1036,
      "step": 80
    },
    {
      "epoch": 0.33986013986013985,
      "grad_norm": 30.54612159729004,
      "learning_rate": 4.05e-06,
      "loss": 0.1044,
      "step": 81
    },
    {
      "epoch": 0.34405594405594403,
      "grad_norm": 23.064117431640625,
      "learning_rate": 4.1e-06,
      "loss": 0.1082,
      "step": 82
    },
    {
      "epoch": 0.34825174825174826,
      "grad_norm": 77.5650863647461,
      "learning_rate": 4.15e-06,
      "loss": 0.1079,
      "step": 83
    },
    {
      "epoch": 0.35244755244755244,
      "grad_norm": 81.39286041259766,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0984,
      "step": 84
    },
    {
      "epoch": 0.35664335664335667,
      "grad_norm": 40.70750045776367,
      "learning_rate": 4.25e-06,
      "loss": 0.1047,
      "step": 85
    },
    {
      "epoch": 0.36083916083916084,
      "grad_norm": 8.890117645263672,
      "learning_rate": 4.3e-06,
      "loss": 0.0978,
      "step": 86
    },
    {
      "epoch": 0.365034965034965,
      "grad_norm": 13.848636627197266,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.101,
      "step": 87
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 21.611827850341797,
      "learning_rate": 4.4e-06,
      "loss": 0.0901,
      "step": 88
    },
    {
      "epoch": 0.3734265734265734,
      "grad_norm": 22.858949661254883,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0958,
      "step": 89
    },
    {
      "epoch": 0.3776223776223776,
      "grad_norm": 3.473940372467041,
      "learning_rate": 4.5e-06,
      "loss": 0.0921,
      "step": 90
    },
    {
      "epoch": 0.38181818181818183,
      "grad_norm": 9.609582901000977,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.1034,
      "step": 91
    },
    {
      "epoch": 0.386013986013986,
      "grad_norm": 8.03201961517334,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0928,
      "step": 92
    },
    {
      "epoch": 0.3902097902097902,
      "grad_norm": 1.6049108505249023,
      "learning_rate": 4.65e-06,
      "loss": 0.0867,
      "step": 93
    },
    {
      "epoch": 0.3944055944055944,
      "grad_norm": 6.763339996337891,
      "learning_rate": 4.7e-06,
      "loss": 0.0911,
      "step": 94
    },
    {
      "epoch": 0.3986013986013986,
      "grad_norm": 7.147759914398193,
      "learning_rate": 4.75e-06,
      "loss": 0.1057,
      "step": 95
    },
    {
      "epoch": 0.4027972027972028,
      "grad_norm": 7.845475673675537,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.1026,
      "step": 96
    },
    {
      "epoch": 0.406993006993007,
      "grad_norm": 30.894298553466797,
      "learning_rate": 4.85e-06,
      "loss": 0.0894,
      "step": 97
    },
    {
      "epoch": 0.4111888111888112,
      "grad_norm": 51.0138053894043,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.1011,
      "step": 98
    },
    {
      "epoch": 0.4153846153846154,
      "grad_norm": 45.050018310546875,
      "learning_rate": 4.95e-06,
      "loss": 0.0943,
      "step": 99
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 9.880765914916992,
      "learning_rate": 5e-06,
      "loss": 0.0923,
      "step": 100
    },
    {
      "epoch": 0.42377622377622376,
      "grad_norm": 1.3705830574035645,
      "learning_rate": 4.999993004586259e-06,
      "loss": 0.0967,
      "step": 101
    },
    {
      "epoch": 0.427972027972028,
      "grad_norm": 4.682475566864014,
      "learning_rate": 4.999972018384183e-06,
      "loss": 0.0916,
      "step": 102
    },
    {
      "epoch": 0.43216783216783217,
      "grad_norm": 3.074434518814087,
      "learning_rate": 4.9999370415112185e-06,
      "loss": 0.0864,
      "step": 103
    },
    {
      "epoch": 0.43636363636363634,
      "grad_norm": 1.248001217842102,
      "learning_rate": 4.999888074163108e-06,
      "loss": 0.0944,
      "step": 104
    },
    {
      "epoch": 0.4405594405594406,
      "grad_norm": 2.949608325958252,
      "learning_rate": 4.999825116613888e-06,
      "loss": 0.0878,
      "step": 105
    },
    {
      "epoch": 0.44475524475524475,
      "grad_norm": 22.338565826416016,
      "learning_rate": 4.999748169215891e-06,
      "loss": 0.0848,
      "step": 106
    },
    {
      "epoch": 0.4489510489510489,
      "grad_norm": 33.91677474975586,
      "learning_rate": 4.999657232399739e-06,
      "loss": 0.0959,
      "step": 107
    },
    {
      "epoch": 0.45314685314685316,
      "grad_norm": 53.919132232666016,
      "learning_rate": 4.999552306674345e-06,
      "loss": 0.0909,
      "step": 108
    },
    {
      "epoch": 0.45734265734265733,
      "grad_norm": 28.08509063720703,
      "learning_rate": 4.999433392626908e-06,
      "loss": 0.0839,
      "step": 109
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 6.412335395812988,
      "learning_rate": 4.999300490922911e-06,
      "loss": 0.0823,
      "step": 110
    },
    {
      "epoch": 0.46573426573426574,
      "grad_norm": 2.714561700820923,
      "learning_rate": 4.999153602306115e-06,
      "loss": 0.0949,
      "step": 111
    },
    {
      "epoch": 0.4699300699300699,
      "grad_norm": 1.5083353519439697,
      "learning_rate": 4.998992727598557e-06,
      "loss": 0.0856,
      "step": 112
    },
    {
      "epoch": 0.47412587412587415,
      "grad_norm": 1.422822117805481,
      "learning_rate": 4.998817867700546e-06,
      "loss": 0.0837,
      "step": 113
    },
    {
      "epoch": 0.4783216783216783,
      "grad_norm": 7.532190799713135,
      "learning_rate": 4.998629023590656e-06,
      "loss": 0.0901,
      "step": 114
    },
    {
      "epoch": 0.4825174825174825,
      "grad_norm": 28.832021713256836,
      "learning_rate": 4.998426196325721e-06,
      "loss": 0.0852,
      "step": 115
    },
    {
      "epoch": 0.48671328671328673,
      "grad_norm": 13.580469131469727,
      "learning_rate": 4.998209387040829e-06,
      "loss": 0.0879,
      "step": 116
    },
    {
      "epoch": 0.4909090909090909,
      "grad_norm": 11.114592552185059,
      "learning_rate": 4.9979785969493165e-06,
      "loss": 0.0819,
      "step": 117
    },
    {
      "epoch": 0.4951048951048951,
      "grad_norm": 1.150378942489624,
      "learning_rate": 4.9977338273427625e-06,
      "loss": 0.0773,
      "step": 118
    },
    {
      "epoch": 0.4993006993006993,
      "grad_norm": 1.427876353263855,
      "learning_rate": 4.997475079590977e-06,
      "loss": 0.0848,
      "step": 119
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 1.115370750427246,
      "learning_rate": 4.9972023551419995e-06,
      "loss": 0.0766,
      "step": 120
    },
    {
      "epoch": 0.5076923076923077,
      "grad_norm": 1.1270095109939575,
      "learning_rate": 4.996915655522086e-06,
      "loss": 0.0833,
      "step": 121
    },
    {
      "epoch": 0.5118881118881119,
      "grad_norm": 1.219711184501648,
      "learning_rate": 4.9966149823357e-06,
      "loss": 0.0881,
      "step": 122
    },
    {
      "epoch": 0.5160839160839161,
      "grad_norm": 0.8859412670135498,
      "learning_rate": 4.996300337265512e-06,
      "loss": 0.0817,
      "step": 123
    },
    {
      "epoch": 0.5202797202797202,
      "grad_norm": 0.5764117240905762,
      "learning_rate": 4.995971722072379e-06,
      "loss": 0.0792,
      "step": 124
    },
    {
      "epoch": 0.5244755244755245,
      "grad_norm": 0.8799321055412292,
      "learning_rate": 4.9956291385953385e-06,
      "loss": 0.0781,
      "step": 125
    },
    {
      "epoch": 0.5286713286713287,
      "grad_norm": 1.0168715715408325,
      "learning_rate": 4.9952725887516015e-06,
      "loss": 0.0797,
      "step": 126
    },
    {
      "epoch": 0.5328671328671328,
      "grad_norm": 1.323622465133667,
      "learning_rate": 4.99490207453654e-06,
      "loss": 0.0842,
      "step": 127
    },
    {
      "epoch": 0.5370629370629371,
      "grad_norm": 1.757987141609192,
      "learning_rate": 4.9945175980236745e-06,
      "loss": 0.0766,
      "step": 128
    },
    {
      "epoch": 0.5412587412587413,
      "grad_norm": 2.3605685234069824,
      "learning_rate": 4.994119161364663e-06,
      "loss": 0.0772,
      "step": 129
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.0868511199951172,
      "learning_rate": 4.993706766789287e-06,
      "loss": 0.075,
      "step": 130
    },
    {
      "epoch": 0.5496503496503496,
      "grad_norm": 5.12212610244751,
      "learning_rate": 4.993280416605444e-06,
      "loss": 0.0778,
      "step": 131
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 13.817222595214844,
      "learning_rate": 4.992840113199131e-06,
      "loss": 0.0776,
      "step": 132
    },
    {
      "epoch": 0.558041958041958,
      "grad_norm": 1.0479037761688232,
      "learning_rate": 4.992385859034431e-06,
      "loss": 0.0833,
      "step": 133
    },
    {
      "epoch": 0.5622377622377622,
      "grad_norm": 4.102362155914307,
      "learning_rate": 4.991917656653501e-06,
      "loss": 0.0789,
      "step": 134
    },
    {
      "epoch": 0.5664335664335665,
      "grad_norm": 3.706054210662842,
      "learning_rate": 4.991435508676557e-06,
      "loss": 0.0703,
      "step": 135
    },
    {
      "epoch": 0.5706293706293706,
      "grad_norm": 0.9211217761039734,
      "learning_rate": 4.990939417801859e-06,
      "loss": 0.0832,
      "step": 136
    },
    {
      "epoch": 0.5748251748251748,
      "grad_norm": 1.108568787574768,
      "learning_rate": 4.990429386805694e-06,
      "loss": 0.0745,
      "step": 137
    },
    {
      "epoch": 0.579020979020979,
      "grad_norm": 0.732056736946106,
      "learning_rate": 4.989905418542366e-06,
      "loss": 0.0745,
      "step": 138
    },
    {
      "epoch": 0.5832167832167832,
      "grad_norm": 0.5949814319610596,
      "learning_rate": 4.989367515944173e-06,
      "loss": 0.0809,
      "step": 139
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 1.0817596912384033,
      "learning_rate": 4.988815682021398e-06,
      "loss": 0.0698,
      "step": 140
    },
    {
      "epoch": 0.5916083916083916,
      "grad_norm": 1.5362236499786377,
      "learning_rate": 4.988249919862285e-06,
      "loss": 0.073,
      "step": 141
    },
    {
      "epoch": 0.5958041958041959,
      "grad_norm": 1.4847773313522339,
      "learning_rate": 4.987670232633027e-06,
      "loss": 0.0765,
      "step": 142
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7822750806808472,
      "learning_rate": 4.987076623577745e-06,
      "loss": 0.0733,
      "step": 143
    },
    {
      "epoch": 0.6041958041958042,
      "grad_norm": 0.9540371894836426,
      "learning_rate": 4.986469096018472e-06,
      "loss": 0.0752,
      "step": 144
    },
    {
      "epoch": 0.6083916083916084,
      "grad_norm": 1.355391263961792,
      "learning_rate": 4.9858476533551335e-06,
      "loss": 0.0707,
      "step": 145
    },
    {
      "epoch": 0.6125874125874126,
      "grad_norm": 0.9062367081642151,
      "learning_rate": 4.985212299065528e-06,
      "loss": 0.0748,
      "step": 146
    },
    {
      "epoch": 0.6167832167832168,
      "grad_norm": 0.7668821811676025,
      "learning_rate": 4.984563036705309e-06,
      "loss": 0.0667,
      "step": 147
    },
    {
      "epoch": 0.620979020979021,
      "grad_norm": 0.7187187671661377,
      "learning_rate": 4.983899869907963e-06,
      "loss": 0.0719,
      "step": 148
    },
    {
      "epoch": 0.6251748251748251,
      "grad_norm": 1.0374109745025635,
      "learning_rate": 4.9832228023847905e-06,
      "loss": 0.0724,
      "step": 149
    },
    {
      "epoch": 0.6293706293706294,
      "grad_norm": 1.6743290424346924,
      "learning_rate": 4.982531837924887e-06,
      "loss": 0.0745,
      "step": 150
    },
    {
      "epoch": 0.6335664335664336,
      "grad_norm": 0.8888541460037231,
      "learning_rate": 4.981826980395117e-06,
      "loss": 0.0683,
      "step": 151
    },
    {
      "epoch": 0.6377622377622377,
      "grad_norm": 0.8322030305862427,
      "learning_rate": 4.981108233740096e-06,
      "loss": 0.066,
      "step": 152
    },
    {
      "epoch": 0.641958041958042,
      "grad_norm": 1.1838620901107788,
      "learning_rate": 4.98037560198217e-06,
      "loss": 0.0815,
      "step": 153
    },
    {
      "epoch": 0.6461538461538462,
      "grad_norm": 1.9783540964126587,
      "learning_rate": 4.979629089221387e-06,
      "loss": 0.0732,
      "step": 154
    },
    {
      "epoch": 0.6503496503496503,
      "grad_norm": 1.0773838758468628,
      "learning_rate": 4.978868699635482e-06,
      "loss": 0.0635,
      "step": 155
    },
    {
      "epoch": 0.6545454545454545,
      "grad_norm": 1.0689246654510498,
      "learning_rate": 4.978094437479843e-06,
      "loss": 0.0705,
      "step": 156
    },
    {
      "epoch": 0.6587412587412588,
      "grad_norm": 0.9070070385932922,
      "learning_rate": 4.977306307087502e-06,
      "loss": 0.0628,
      "step": 157
    },
    {
      "epoch": 0.6629370629370629,
      "grad_norm": 2.045474052429199,
      "learning_rate": 4.976504312869093e-06,
      "loss": 0.0645,
      "step": 158
    },
    {
      "epoch": 0.6671328671328671,
      "grad_norm": 4.762063503265381,
      "learning_rate": 4.975688459312844e-06,
      "loss": 0.0661,
      "step": 159
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 10.791966438293457,
      "learning_rate": 4.97485875098454e-06,
      "loss": 0.0657,
      "step": 160
    },
    {
      "epoch": 0.6755244755244755,
      "grad_norm": 1.7596170902252197,
      "learning_rate": 4.974015192527505e-06,
      "loss": 0.075,
      "step": 161
    },
    {
      "epoch": 0.6797202797202797,
      "grad_norm": 2.43743896484375,
      "learning_rate": 4.97315778866257e-06,
      "loss": 0.0622,
      "step": 162
    },
    {
      "epoch": 0.6839160839160839,
      "grad_norm": 1.6388510465621948,
      "learning_rate": 4.972286544188052e-06,
      "loss": 0.0601,
      "step": 163
    },
    {
      "epoch": 0.6881118881118881,
      "grad_norm": 0.8130991458892822,
      "learning_rate": 4.971401463979722e-06,
      "loss": 0.0607,
      "step": 164
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.8717385530471802,
      "learning_rate": 4.970502552990782e-06,
      "loss": 0.0706,
      "step": 165
    },
    {
      "epoch": 0.6965034965034965,
      "grad_norm": 0.8151756525039673,
      "learning_rate": 4.969589816251837e-06,
      "loss": 0.0628,
      "step": 166
    },
    {
      "epoch": 0.7006993006993008,
      "grad_norm": 0.7532489895820618,
      "learning_rate": 4.968663258870862e-06,
      "loss": 0.0687,
      "step": 167
    },
    {
      "epoch": 0.7048951048951049,
      "grad_norm": 0.5839558839797974,
      "learning_rate": 4.967722886033181e-06,
      "loss": 0.0554,
      "step": 168
    },
    {
      "epoch": 0.7090909090909091,
      "grad_norm": 0.6098427772521973,
      "learning_rate": 4.966768703001429e-06,
      "loss": 0.0639,
      "step": 169
    },
    {
      "epoch": 0.7132867132867133,
      "grad_norm": 0.9210315942764282,
      "learning_rate": 4.965800715115531e-06,
      "loss": 0.0704,
      "step": 170
    },
    {
      "epoch": 0.7174825174825175,
      "grad_norm": 0.9615612626075745,
      "learning_rate": 4.964818927792669e-06,
      "loss": 0.068,
      "step": 171
    },
    {
      "epoch": 0.7216783216783217,
      "grad_norm": 0.7204346060752869,
      "learning_rate": 4.963823346527249e-06,
      "loss": 0.0663,
      "step": 172
    },
    {
      "epoch": 0.7258741258741259,
      "grad_norm": 0.7109172344207764,
      "learning_rate": 4.962813976890872e-06,
      "loss": 0.0603,
      "step": 173
    },
    {
      "epoch": 0.73006993006993,
      "grad_norm": 0.9478862881660461,
      "learning_rate": 4.961790824532306e-06,
      "loss": 0.061,
      "step": 174
    },
    {
      "epoch": 0.7342657342657343,
      "grad_norm": 0.7901272773742676,
      "learning_rate": 4.96075389517745e-06,
      "loss": 0.0572,
      "step": 175
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 0.7225059270858765,
      "learning_rate": 4.959703194629304e-06,
      "loss": 0.0567,
      "step": 176
    },
    {
      "epoch": 0.7426573426573426,
      "grad_norm": 0.693332850933075,
      "learning_rate": 4.958638728767936e-06,
      "loss": 0.0545,
      "step": 177
    },
    {
      "epoch": 0.7468531468531469,
      "grad_norm": 0.834145724773407,
      "learning_rate": 4.95756050355045e-06,
      "loss": 0.0671,
      "step": 178
    },
    {
      "epoch": 0.7510489510489511,
      "grad_norm": 1.00062894821167,
      "learning_rate": 4.956468525010948e-06,
      "loss": 0.0539,
      "step": 179
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 1.1205754280090332,
      "learning_rate": 4.955362799260507e-06,
      "loss": 0.0529,
      "step": 180
    },
    {
      "epoch": 0.7594405594405594,
      "grad_norm": 2.0870769023895264,
      "learning_rate": 4.954243332487133e-06,
      "loss": 0.059,
      "step": 181
    },
    {
      "epoch": 0.7636363636363637,
      "grad_norm": 2.168609619140625,
      "learning_rate": 4.953110130955733e-06,
      "loss": 0.0534,
      "step": 182
    },
    {
      "epoch": 0.7678321678321678,
      "grad_norm": 1.5437854528427124,
      "learning_rate": 4.9519632010080765e-06,
      "loss": 0.0472,
      "step": 183
    },
    {
      "epoch": 0.772027972027972,
      "grad_norm": 2.0716335773468018,
      "learning_rate": 4.950802549062764e-06,
      "loss": 0.0605,
      "step": 184
    },
    {
      "epoch": 0.7762237762237763,
      "grad_norm": 1.5331146717071533,
      "learning_rate": 4.9496281816151895e-06,
      "loss": 0.0615,
      "step": 185
    },
    {
      "epoch": 0.7804195804195804,
      "grad_norm": 0.8838733434677124,
      "learning_rate": 4.948440105237499e-06,
      "loss": 0.0494,
      "step": 186
    },
    {
      "epoch": 0.7846153846153846,
      "grad_norm": 0.8043892979621887,
      "learning_rate": 4.947238326578563e-06,
      "loss": 0.0568,
      "step": 187
    },
    {
      "epoch": 0.7888111888111888,
      "grad_norm": 1.4767056703567505,
      "learning_rate": 4.946022852363932e-06,
      "loss": 0.0565,
      "step": 188
    },
    {
      "epoch": 0.793006993006993,
      "grad_norm": 1.083964467048645,
      "learning_rate": 4.944793689395802e-06,
      "loss": 0.0582,
      "step": 189
    },
    {
      "epoch": 0.7972027972027972,
      "grad_norm": 0.99649977684021,
      "learning_rate": 4.943550844552978e-06,
      "loss": 0.048,
      "step": 190
    },
    {
      "epoch": 0.8013986013986014,
      "grad_norm": 0.674566924571991,
      "learning_rate": 4.9422943247908275e-06,
      "loss": 0.0485,
      "step": 191
    },
    {
      "epoch": 0.8055944055944056,
      "grad_norm": 1.1796026229858398,
      "learning_rate": 4.9410241371412525e-06,
      "loss": 0.0501,
      "step": 192
    },
    {
      "epoch": 0.8097902097902098,
      "grad_norm": 0.9794552326202393,
      "learning_rate": 4.939740288712644e-06,
      "loss": 0.0527,
      "step": 193
    },
    {
      "epoch": 0.813986013986014,
      "grad_norm": 0.8760250210762024,
      "learning_rate": 4.938442786689843e-06,
      "loss": 0.0539,
      "step": 194
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 1.3841289281845093,
      "learning_rate": 4.9371316383341e-06,
      "loss": 0.0543,
      "step": 195
    },
    {
      "epoch": 0.8223776223776224,
      "grad_norm": 1.4383552074432373,
      "learning_rate": 4.935806850983034e-06,
      "loss": 0.0479,
      "step": 196
    },
    {
      "epoch": 0.8265734265734266,
      "grad_norm": 1.7970709800720215,
      "learning_rate": 4.934468432050594e-06,
      "loss": 0.049,
      "step": 197
    },
    {
      "epoch": 0.8307692307692308,
      "grad_norm": 1.0703388452529907,
      "learning_rate": 4.933116389027017e-06,
      "loss": 0.0527,
      "step": 198
    },
    {
      "epoch": 0.8349650349650349,
      "grad_norm": 3.151549816131592,
      "learning_rate": 4.931750729478781e-06,
      "loss": 0.0529,
      "step": 199
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 1.772434949874878,
      "learning_rate": 4.9303714610485705e-06,
      "loss": 0.0535,
      "step": 200
    },
    {
      "epoch": 0.8433566433566434,
      "grad_norm": 0.8801355361938477,
      "learning_rate": 4.928978591455228e-06,
      "loss": 0.0504,
      "step": 201
    },
    {
      "epoch": 0.8475524475524475,
      "grad_norm": 1.4067658185958862,
      "learning_rate": 4.9275721284937115e-06,
      "loss": 0.049,
      "step": 202
    },
    {
      "epoch": 0.8517482517482518,
      "grad_norm": 1.0001884698867798,
      "learning_rate": 4.926152080035055e-06,
      "loss": 0.0614,
      "step": 203
    },
    {
      "epoch": 0.855944055944056,
      "grad_norm": 1.1527674198150635,
      "learning_rate": 4.924718454026318e-06,
      "loss": 0.0479,
      "step": 204
    },
    {
      "epoch": 0.8601398601398601,
      "grad_norm": 1.1766809225082397,
      "learning_rate": 4.923271258490547e-06,
      "loss": 0.0482,
      "step": 205
    },
    {
      "epoch": 0.8643356643356643,
      "grad_norm": 0.8541592359542847,
      "learning_rate": 4.921810501526728e-06,
      "loss": 0.0458,
      "step": 206
    },
    {
      "epoch": 0.8685314685314686,
      "grad_norm": 1.3551443815231323,
      "learning_rate": 4.9203361913097394e-06,
      "loss": 0.0497,
      "step": 207
    },
    {
      "epoch": 0.8727272727272727,
      "grad_norm": 1.417004108428955,
      "learning_rate": 4.918848336090309e-06,
      "loss": 0.0472,
      "step": 208
    },
    {
      "epoch": 0.8769230769230769,
      "grad_norm": 1.3648598194122314,
      "learning_rate": 4.917346944194969e-06,
      "loss": 0.0415,
      "step": 209
    },
    {
      "epoch": 0.8811188811188811,
      "grad_norm": 1.0727978944778442,
      "learning_rate": 4.915832024026002e-06,
      "loss": 0.0421,
      "step": 210
    },
    {
      "epoch": 0.8853146853146853,
      "grad_norm": 2.974808692932129,
      "learning_rate": 4.914303584061405e-06,
      "loss": 0.0474,
      "step": 211
    },
    {
      "epoch": 0.8895104895104895,
      "grad_norm": 1.6205039024353027,
      "learning_rate": 4.912761632854834e-06,
      "loss": 0.0445,
      "step": 212
    },
    {
      "epoch": 0.8937062937062937,
      "grad_norm": 4.179663181304932,
      "learning_rate": 4.911206179035557e-06,
      "loss": 0.0438,
      "step": 213
    },
    {
      "epoch": 0.8979020979020979,
      "grad_norm": 10.701528549194336,
      "learning_rate": 4.90963723130841e-06,
      "loss": 0.0399,
      "step": 214
    },
    {
      "epoch": 0.9020979020979021,
      "grad_norm": 11.126282691955566,
      "learning_rate": 4.908054798453744e-06,
      "loss": 0.0499,
      "step": 215
    },
    {
      "epoch": 0.9062937062937063,
      "grad_norm": 8.534737586975098,
      "learning_rate": 4.906458889327375e-06,
      "loss": 0.0445,
      "step": 216
    },
    {
      "epoch": 0.9104895104895104,
      "grad_norm": 1.3229663372039795,
      "learning_rate": 4.90484951286054e-06,
      "loss": 0.0495,
      "step": 217
    },
    {
      "epoch": 0.9146853146853147,
      "grad_norm": 1.9636162519454956,
      "learning_rate": 4.903226678059842e-06,
      "loss": 0.0464,
      "step": 218
    },
    {
      "epoch": 0.9188811188811189,
      "grad_norm": 1.5941684246063232,
      "learning_rate": 4.901590394007202e-06,
      "loss": 0.0469,
      "step": 219
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 2.0621442794799805,
      "learning_rate": 4.899940669859807e-06,
      "loss": 0.0438,
      "step": 220
    },
    {
      "epoch": 0.9272727272727272,
      "grad_norm": 0.8177353739738464,
      "learning_rate": 4.89827751485006e-06,
      "loss": 0.0519,
      "step": 221
    },
    {
      "epoch": 0.9314685314685315,
      "grad_norm": 6.465226173400879,
      "learning_rate": 4.896600938285526e-06,
      "loss": 0.0424,
      "step": 222
    },
    {
      "epoch": 0.9356643356643357,
      "grad_norm": 2.424146890640259,
      "learning_rate": 4.8949109495488834e-06,
      "loss": 0.0444,
      "step": 223
    },
    {
      "epoch": 0.9398601398601398,
      "grad_norm": 3.0828588008880615,
      "learning_rate": 4.893207558097867e-06,
      "loss": 0.0402,
      "step": 224
    },
    {
      "epoch": 0.9440559440559441,
      "grad_norm": 0.6914602518081665,
      "learning_rate": 4.891490773465221e-06,
      "loss": 0.038,
      "step": 225
    },
    {
      "epoch": 0.9482517482517483,
      "grad_norm": 0.7816252112388611,
      "learning_rate": 4.8897606052586384e-06,
      "loss": 0.0419,
      "step": 226
    },
    {
      "epoch": 0.9524475524475524,
      "grad_norm": 1.139478325843811,
      "learning_rate": 4.8880170631607145e-06,
      "loss": 0.0402,
      "step": 227
    },
    {
      "epoch": 0.9566433566433566,
      "grad_norm": 0.9213783740997314,
      "learning_rate": 4.8862601569288885e-06,
      "loss": 0.0413,
      "step": 228
    },
    {
      "epoch": 0.9608391608391609,
      "grad_norm": 1.3993144035339355,
      "learning_rate": 4.8844898963953875e-06,
      "loss": 0.0383,
      "step": 229
    },
    {
      "epoch": 0.965034965034965,
      "grad_norm": 0.8406299352645874,
      "learning_rate": 4.8827062914671775e-06,
      "loss": 0.0345,
      "step": 230
    },
    {
      "epoch": 0.9692307692307692,
      "grad_norm": 1.1061815023422241,
      "learning_rate": 4.880909352125901e-06,
      "loss": 0.0411,
      "step": 231
    },
    {
      "epoch": 0.9734265734265735,
      "grad_norm": 1.6662583351135254,
      "learning_rate": 4.879099088427824e-06,
      "loss": 0.0376,
      "step": 232
    },
    {
      "epoch": 0.9776223776223776,
      "grad_norm": 1.7703438997268677,
      "learning_rate": 4.877275510503783e-06,
      "loss": 0.0412,
      "step": 233
    },
    {
      "epoch": 0.9818181818181818,
      "grad_norm": 0.8591011762619019,
      "learning_rate": 4.875438628559124e-06,
      "loss": 0.038,
      "step": 234
    },
    {
      "epoch": 0.986013986013986,
      "grad_norm": 1.0391316413879395,
      "learning_rate": 4.8735884528736444e-06,
      "loss": 0.0446,
      "step": 235
    },
    {
      "epoch": 0.9902097902097902,
      "grad_norm": 1.1077414751052856,
      "learning_rate": 4.871724993801541e-06,
      "loss": 0.037,
      "step": 236
    },
    {
      "epoch": 0.9944055944055944,
      "grad_norm": 1.3140100240707397,
      "learning_rate": 4.869848261771348e-06,
      "loss": 0.0379,
      "step": 237
    },
    {
      "epoch": 0.9986013986013986,
      "grad_norm": 0.7918304204940796,
      "learning_rate": 4.867958267285879e-06,
      "loss": 0.037,
      "step": 238
    }
  ],
  "logging_steps": 1,
  "max_steps": 1428,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 238,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.011967338289562e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}