{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9944055944055945,
  "eval_steps": 500,
  "global_step": 476,
  "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
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7918304204940796,
      "learning_rate": 4.866055020922167e-06,
      "loss": 0.0281,
      "step": 239
    },
    {
      "epoch": 1.0041958041958041,
      "grad_norm": 1.9362425804138184,
      "learning_rate": 4.864138533331411e-06,
      "loss": 0.0259,
      "step": 240
    },
    {
      "epoch": 1.0083916083916085,
      "grad_norm": 1.20184326171875,
      "learning_rate": 4.862208815238908e-06,
      "loss": 0.0214,
      "step": 241
    },
    {
      "epoch": 1.0125874125874126,
      "grad_norm": 1.2968955039978027,
      "learning_rate": 4.860265877444001e-06,
      "loss": 0.0275,
      "step": 242
    },
    {
      "epoch": 1.0167832167832167,
      "grad_norm": 0.7607834935188293,
      "learning_rate": 4.85830973082001e-06,
      "loss": 0.0263,
      "step": 243
    },
    {
      "epoch": 1.020979020979021,
      "grad_norm": 0.9456799626350403,
      "learning_rate": 4.8563403863141825e-06,
      "loss": 0.0276,
      "step": 244
    },
    {
      "epoch": 1.0251748251748252,
      "grad_norm": 0.7519564032554626,
      "learning_rate": 4.854357854947619e-06,
      "loss": 0.0245,
      "step": 245
    },
    {
      "epoch": 1.0293706293706293,
      "grad_norm": 1.3622503280639648,
      "learning_rate": 4.852362147815225e-06,
      "loss": 0.0204,
      "step": 246
    },
    {
      "epoch": 1.0335664335664336,
      "grad_norm": 1.1832894086837769,
      "learning_rate": 4.850353276085634e-06,
      "loss": 0.0269,
      "step": 247
    },
    {
      "epoch": 1.0377622377622377,
      "grad_norm": 1.7349433898925781,
      "learning_rate": 4.84833125100116e-06,
      "loss": 0.0247,
      "step": 248
    },
    {
      "epoch": 1.0419580419580419,
      "grad_norm": 0.978853166103363,
      "learning_rate": 4.8462960838777234e-06,
      "loss": 0.0278,
      "step": 249
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 3.0965521335601807,
      "learning_rate": 4.844247786104794e-06,
      "loss": 0.0326,
      "step": 250
    },
    {
      "epoch": 1.0503496503496503,
      "grad_norm": 12.701726913452148,
      "learning_rate": 4.842186369145323e-06,
      "loss": 0.0275,
      "step": 251
    },
    {
      "epoch": 1.0545454545454545,
      "grad_norm": 1.9011986255645752,
      "learning_rate": 4.840111844535682e-06,
      "loss": 0.026,
      "step": 252
    },
    {
      "epoch": 1.0587412587412588,
      "grad_norm": 1.687672734260559,
      "learning_rate": 4.838024223885599e-06,
      "loss": 0.0247,
      "step": 253
    },
    {
      "epoch": 1.062937062937063,
      "grad_norm": 1.5581176280975342,
      "learning_rate": 4.835923518878088e-06,
      "loss": 0.0174,
      "step": 254
    },
    {
      "epoch": 1.067132867132867,
      "grad_norm": 0.980636715888977,
      "learning_rate": 4.833809741269392e-06,
      "loss": 0.0294,
      "step": 255
    },
    {
      "epoch": 1.0713286713286714,
      "grad_norm": 1.0523900985717773,
      "learning_rate": 4.8316829028889076e-06,
      "loss": 0.0241,
      "step": 256
    },
    {
      "epoch": 1.0755244755244755,
      "grad_norm": 1.6697787046432495,
      "learning_rate": 4.829543015639128e-06,
      "loss": 0.022,
      "step": 257
    },
    {
      "epoch": 1.0797202797202796,
      "grad_norm": 0.9570637941360474,
      "learning_rate": 4.82739009149557e-06,
      "loss": 0.0251,
      "step": 258
    },
    {
      "epoch": 1.083916083916084,
      "grad_norm": 1.8936893939971924,
      "learning_rate": 4.825224142506708e-06,
      "loss": 0.0229,
      "step": 259
    },
    {
      "epoch": 1.088111888111888,
      "grad_norm": 1.288090467453003,
      "learning_rate": 4.823045180793914e-06,
      "loss": 0.0258,
      "step": 260
    },
    {
      "epoch": 1.0923076923076924,
      "grad_norm": 1.9112778902053833,
      "learning_rate": 4.820853218551375e-06,
      "loss": 0.0206,
      "step": 261
    },
    {
      "epoch": 1.0965034965034965,
      "grad_norm": 1.1951854228973389,
      "learning_rate": 4.818648268046038e-06,
      "loss": 0.0201,
      "step": 262
    },
    {
      "epoch": 1.1006993006993007,
      "grad_norm": 12.678963661193848,
      "learning_rate": 4.816430341617536e-06,
      "loss": 0.021,
      "step": 263
    },
    {
      "epoch": 1.104895104895105,
      "grad_norm": 11.643608093261719,
      "learning_rate": 4.8141994516781196e-06,
      "loss": 0.0208,
      "step": 264
    },
    {
      "epoch": 1.1090909090909091,
      "grad_norm": 9.084321022033691,
      "learning_rate": 4.811955610712588e-06,
      "loss": 0.0236,
      "step": 265
    },
    {
      "epoch": 1.1132867132867132,
      "grad_norm": 1.2543950080871582,
      "learning_rate": 4.809698831278217e-06,
      "loss": 0.0224,
      "step": 266
    },
    {
      "epoch": 1.1174825174825176,
      "grad_norm": 1.0576941967010498,
      "learning_rate": 4.8074291260046925e-06,
      "loss": 0.0225,
      "step": 267
    },
    {
      "epoch": 1.1216783216783217,
      "grad_norm": 0.9116222262382507,
      "learning_rate": 4.805146507594034e-06,
      "loss": 0.0206,
      "step": 268
    },
    {
      "epoch": 1.1258741258741258,
      "grad_norm": 1.11386239528656,
      "learning_rate": 4.8028509888205325e-06,
      "loss": 0.0183,
      "step": 269
    },
    {
      "epoch": 1.1300699300699302,
      "grad_norm": 0.7633559703826904,
      "learning_rate": 4.800542582530668e-06,
      "loss": 0.0226,
      "step": 270
    },
    {
      "epoch": 1.1342657342657343,
      "grad_norm": 1.9914929866790771,
      "learning_rate": 4.79822130164305e-06,
      "loss": 0.0231,
      "step": 271
    },
    {
      "epoch": 1.1384615384615384,
      "grad_norm": 0.8240312337875366,
      "learning_rate": 4.7958871591483305e-06,
      "loss": 0.0158,
      "step": 272
    },
    {
      "epoch": 1.1426573426573428,
      "grad_norm": 0.8062412738800049,
      "learning_rate": 4.793540168109146e-06,
      "loss": 0.0165,
      "step": 273
    },
    {
      "epoch": 1.1468531468531469,
      "grad_norm": 0.8970699906349182,
      "learning_rate": 4.791180341660035e-06,
      "loss": 0.0223,
      "step": 274
    },
    {
      "epoch": 1.151048951048951,
      "grad_norm": 1.1959490776062012,
      "learning_rate": 4.788807693007367e-06,
      "loss": 0.0185,
      "step": 275
    },
    {
      "epoch": 1.1552447552447553,
      "grad_norm": 1.3478264808654785,
      "learning_rate": 4.786422235429269e-06,
      "loss": 0.0219,
      "step": 276
    },
    {
      "epoch": 1.1594405594405595,
      "grad_norm": 2.527574300765991,
      "learning_rate": 4.784023982275551e-06,
      "loss": 0.0171,
      "step": 277
    },
    {
      "epoch": 1.1636363636363636,
      "grad_norm": 2.0560972690582275,
      "learning_rate": 4.781612946967632e-06,
      "loss": 0.0232,
      "step": 278
    },
    {
      "epoch": 1.167832167832168,
      "grad_norm": 1.2283650636672974,
      "learning_rate": 4.779189142998463e-06,
      "loss": 0.0228,
      "step": 279
    },
    {
      "epoch": 1.172027972027972,
      "grad_norm": 1.0990374088287354,
      "learning_rate": 4.776752583932455e-06,
      "loss": 0.0207,
      "step": 280
    },
    {
      "epoch": 1.1762237762237762,
      "grad_norm": 0.762873113155365,
      "learning_rate": 4.774303283405396e-06,
      "loss": 0.0192,
      "step": 281
    },
    {
      "epoch": 1.1804195804195805,
      "grad_norm": 3.5389952659606934,
      "learning_rate": 4.771841255124385e-06,
      "loss": 0.0193,
      "step": 282
    },
    {
      "epoch": 1.1846153846153846,
      "grad_norm": 1.5018829107284546,
      "learning_rate": 4.769366512867747e-06,
      "loss": 0.0177,
      "step": 283
    },
    {
      "epoch": 1.1888111888111887,
      "grad_norm": 0.9355039596557617,
      "learning_rate": 4.766879070484957e-06,
      "loss": 0.0222,
      "step": 284
    },
    {
      "epoch": 1.193006993006993,
      "grad_norm": 0.9899689555168152,
      "learning_rate": 4.764378941896568e-06,
      "loss": 0.0237,
      "step": 285
    },
    {
      "epoch": 1.1972027972027972,
      "grad_norm": 0.7639883160591125,
      "learning_rate": 4.761866141094126e-06,
      "loss": 0.0206,
      "step": 286
    },
    {
      "epoch": 1.2013986013986013,
      "grad_norm": 0.9030625224113464,
      "learning_rate": 4.759340682140095e-06,
      "loss": 0.0147,
      "step": 287
    },
    {
      "epoch": 1.2055944055944057,
      "grad_norm": 1.505378246307373,
      "learning_rate": 4.756802579167781e-06,
      "loss": 0.0206,
      "step": 288
    },
    {
      "epoch": 1.2097902097902098,
      "grad_norm": 0.5766127109527588,
      "learning_rate": 4.754251846381248e-06,
      "loss": 0.0138,
      "step": 289
    },
    {
      "epoch": 1.213986013986014,
      "grad_norm": 0.937792181968689,
      "learning_rate": 4.75168849805524e-06,
      "loss": 0.0191,
      "step": 290
    },
    {
      "epoch": 1.2181818181818183,
      "grad_norm": 0.8431217074394226,
      "learning_rate": 4.749112548535104e-06,
      "loss": 0.0192,
      "step": 291
    },
    {
      "epoch": 1.2223776223776224,
      "grad_norm": 0.7842226624488831,
      "learning_rate": 4.746524012236706e-06,
      "loss": 0.0151,
      "step": 292
    },
    {
      "epoch": 1.2265734265734265,
      "grad_norm": 0.9969326853752136,
      "learning_rate": 4.743922903646351e-06,
      "loss": 0.0171,
      "step": 293
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 1.2452991008758545,
      "learning_rate": 4.741309237320703e-06,
      "loss": 0.0203,
      "step": 294
    },
    {
      "epoch": 1.234965034965035,
      "grad_norm": 0.9601305723190308,
      "learning_rate": 4.738683027886706e-06,
      "loss": 0.0144,
      "step": 295
    },
    {
      "epoch": 1.239160839160839,
      "grad_norm": 1.095999002456665,
      "learning_rate": 4.736044290041496e-06,
      "loss": 0.0136,
      "step": 296
    },
    {
      "epoch": 1.2433566433566434,
      "grad_norm": 0.9043452143669128,
      "learning_rate": 4.733393038552323e-06,
      "loss": 0.0216,
      "step": 297
    },
    {
      "epoch": 1.2475524475524475,
      "grad_norm": 1.330333948135376,
      "learning_rate": 4.730729288256468e-06,
      "loss": 0.0186,
      "step": 298
    },
    {
      "epoch": 1.2517482517482517,
      "grad_norm": 12.428945541381836,
      "learning_rate": 4.72805305406116e-06,
      "loss": 0.0224,
      "step": 299
    },
    {
      "epoch": 1.255944055944056,
      "grad_norm": 4.790683746337891,
      "learning_rate": 4.725364350943492e-06,
      "loss": 0.0165,
      "step": 300
    },
    {
      "epoch": 1.2601398601398601,
      "grad_norm": 4.931190490722656,
      "learning_rate": 4.722663193950336e-06,
      "loss": 0.0177,
      "step": 301
    },
    {
      "epoch": 1.2643356643356642,
      "grad_norm": 1.1879394054412842,
      "learning_rate": 4.719949598198258e-06,
      "loss": 0.0167,
      "step": 302
    },
    {
      "epoch": 1.2685314685314686,
      "grad_norm": 0.9889580607414246,
      "learning_rate": 4.717223578873442e-06,
      "loss": 0.0154,
      "step": 303
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.7459793090820312,
      "learning_rate": 4.714485151231593e-06,
      "loss": 0.0115,
      "step": 304
    },
    {
      "epoch": 1.2769230769230768,
      "grad_norm": 1.0093220472335815,
      "learning_rate": 4.711734330597857e-06,
      "loss": 0.014,
      "step": 305
    },
    {
      "epoch": 1.2811188811188812,
      "grad_norm": 0.701112687587738,
      "learning_rate": 4.708971132366739e-06,
      "loss": 0.017,
      "step": 306
    },
    {
      "epoch": 1.2853146853146853,
      "grad_norm": 1.1203210353851318,
      "learning_rate": 4.7061955720020096e-06,
      "loss": 0.0142,
      "step": 307
    },
    {
      "epoch": 1.2895104895104894,
      "grad_norm": 1.1437033414840698,
      "learning_rate": 4.703407665036622e-06,
      "loss": 0.0213,
      "step": 308
    },
    {
      "epoch": 1.2937062937062938,
      "grad_norm": 1.3569507598876953,
      "learning_rate": 4.70060742707263e-06,
      "loss": 0.0127,
      "step": 309
    },
    {
      "epoch": 1.2979020979020979,
      "grad_norm": 0.8850122690200806,
      "learning_rate": 4.697794873781089e-06,
      "loss": 0.0162,
      "step": 310
    },
    {
      "epoch": 1.302097902097902,
      "grad_norm": 0.7463717460632324,
      "learning_rate": 4.69497002090198e-06,
      "loss": 0.0063,
      "step": 311
    },
    {
      "epoch": 1.3062937062937063,
      "grad_norm": 1.432754397392273,
      "learning_rate": 4.692132884244113e-06,
      "loss": 0.0144,
      "step": 312
    },
    {
      "epoch": 1.3104895104895105,
      "grad_norm": 1.0831643342971802,
      "learning_rate": 4.689283479685045e-06,
      "loss": 0.009,
      "step": 313
    },
    {
      "epoch": 1.3146853146853146,
      "grad_norm": 0.908507227897644,
      "learning_rate": 4.686421823170987e-06,
      "loss": 0.0105,
      "step": 314
    },
    {
      "epoch": 1.318881118881119,
      "grad_norm": 0.829806923866272,
      "learning_rate": 4.683547930716716e-06,
      "loss": 0.0132,
      "step": 315
    },
    {
      "epoch": 1.323076923076923,
      "grad_norm": 0.7971309423446655,
      "learning_rate": 4.680661818405485e-06,
      "loss": 0.0129,
      "step": 316
    },
    {
      "epoch": 1.3272727272727272,
      "grad_norm": 1.5599275827407837,
      "learning_rate": 4.677763502388934e-06,
      "loss": 0.0139,
      "step": 317
    },
    {
      "epoch": 1.3314685314685315,
      "grad_norm": 0.7332860231399536,
      "learning_rate": 4.674852998886998e-06,
      "loss": 0.0104,
      "step": 318
    },
    {
      "epoch": 1.3356643356643356,
      "grad_norm": 2.5700080394744873,
      "learning_rate": 4.67193032418782e-06,
      "loss": 0.0132,
      "step": 319
    },
    {
      "epoch": 1.3398601398601397,
      "grad_norm": 1.2639752626419067,
      "learning_rate": 4.668995494647653e-06,
      "loss": 0.0119,
      "step": 320
    },
    {
      "epoch": 1.344055944055944,
      "grad_norm": 0.846431314945221,
      "learning_rate": 4.666048526690775e-06,
      "loss": 0.0135,
      "step": 321
    },
    {
      "epoch": 1.3482517482517482,
      "grad_norm": 0.9121400713920593,
      "learning_rate": 4.663089436809395e-06,
      "loss": 0.015,
      "step": 322
    },
    {
      "epoch": 1.3524475524475523,
      "grad_norm": 0.9713990092277527,
      "learning_rate": 4.660118241563558e-06,
      "loss": 0.0139,
      "step": 323
    },
    {
      "epoch": 1.3566433566433567,
      "grad_norm": 1.3189496994018555,
      "learning_rate": 4.657134957581057e-06,
      "loss": 0.0085,
      "step": 324
    },
    {
      "epoch": 1.3608391608391608,
      "grad_norm": 1.2847219705581665,
      "learning_rate": 4.654139601557337e-06,
      "loss": 0.0089,
      "step": 325
    },
    {
      "epoch": 1.365034965034965,
      "grad_norm": 0.8150692582130432,
      "learning_rate": 4.651132190255401e-06,
      "loss": 0.0081,
      "step": 326
    },
    {
      "epoch": 1.3692307692307693,
      "grad_norm": 1.006005048751831,
      "learning_rate": 4.648112740505718e-06,
      "loss": 0.0147,
      "step": 327
    },
    {
      "epoch": 1.3734265734265734,
      "grad_norm": 1.7578229904174805,
      "learning_rate": 4.645081269206128e-06,
      "loss": 0.0153,
      "step": 328
    },
    {
      "epoch": 1.3776223776223775,
      "grad_norm": 1.0592750310897827,
      "learning_rate": 4.642037793321749e-06,
      "loss": 0.014,
      "step": 329
    },
    {
      "epoch": 1.3818181818181818,
      "grad_norm": 0.9236828684806824,
      "learning_rate": 4.638982329884878e-06,
      "loss": 0.0117,
      "step": 330
    },
    {
      "epoch": 1.386013986013986,
      "grad_norm": 1.0400675535202026,
      "learning_rate": 4.635914895994901e-06,
      "loss": 0.0096,
      "step": 331
    },
    {
      "epoch": 1.39020979020979,
      "grad_norm": 1.1736772060394287,
      "learning_rate": 4.632835508818192e-06,
      "loss": 0.0121,
      "step": 332
    },
    {
      "epoch": 1.3944055944055944,
      "grad_norm": 1.1377946138381958,
      "learning_rate": 4.629744185588022e-06,
      "loss": 0.0102,
      "step": 333
    },
    {
      "epoch": 1.3986013986013985,
      "grad_norm": 0.7523559927940369,
      "learning_rate": 4.626640943604459e-06,
      "loss": 0.0095,
      "step": 334
    },
    {
      "epoch": 1.4027972027972029,
      "grad_norm": 0.6026727557182312,
      "learning_rate": 4.623525800234272e-06,
      "loss": 0.0088,
      "step": 335
    },
    {
      "epoch": 1.406993006993007,
      "grad_norm": 1.142449975013733,
      "learning_rate": 4.620398772910833e-06,
      "loss": 0.0095,
      "step": 336
    },
    {
      "epoch": 1.4111888111888111,
      "grad_norm": 0.7902323007583618,
      "learning_rate": 4.617259879134025e-06,
      "loss": 0.008,
      "step": 337
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 1.2716599702835083,
      "learning_rate": 4.614109136470133e-06,
      "loss": 0.0128,
      "step": 338
    },
    {
      "epoch": 1.4195804195804196,
      "grad_norm": 1.1127336025238037,
      "learning_rate": 4.610946562551759e-06,
      "loss": 0.0152,
      "step": 339
    },
    {
      "epoch": 1.4237762237762237,
      "grad_norm": 0.9858589172363281,
      "learning_rate": 4.607772175077712e-06,
      "loss": 0.0159,
      "step": 340
    },
    {
      "epoch": 1.427972027972028,
      "grad_norm": 1.2081674337387085,
      "learning_rate": 4.604585991812914e-06,
      "loss": 0.0094,
      "step": 341
    },
    {
      "epoch": 1.4321678321678322,
      "grad_norm": 1.1321561336517334,
      "learning_rate": 4.601388030588303e-06,
      "loss": 0.0118,
      "step": 342
    },
    {
      "epoch": 1.4363636363636363,
      "grad_norm": 0.6680831909179688,
      "learning_rate": 4.5981783093007275e-06,
      "loss": 0.0072,
      "step": 343
    },
    {
      "epoch": 1.4405594405594406,
      "grad_norm": 0.9412718415260315,
      "learning_rate": 4.59495684591285e-06,
      "loss": 0.0141,
      "step": 344
    },
    {
      "epoch": 1.4447552447552447,
      "grad_norm": 8.079177856445312,
      "learning_rate": 4.591723658453047e-06,
      "loss": 0.0097,
      "step": 345
    },
    {
      "epoch": 1.4489510489510489,
      "grad_norm": 8.037176132202148,
      "learning_rate": 4.588478765015304e-06,
      "loss": 0.0079,
      "step": 346
    },
    {
      "epoch": 1.4531468531468532,
      "grad_norm": 10.418330192565918,
      "learning_rate": 4.585222183759121e-06,
      "loss": 0.0114,
      "step": 347
    },
    {
      "epoch": 1.4573426573426573,
      "grad_norm": 6.077114582061768,
      "learning_rate": 4.581953932909403e-06,
      "loss": 0.0134,
      "step": 348
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.7302384972572327,
      "learning_rate": 4.578674030756364e-06,
      "loss": 0.0076,
      "step": 349
    },
    {
      "epoch": 1.4657342657342658,
      "grad_norm": 0.8345720171928406,
      "learning_rate": 4.575382495655421e-06,
      "loss": 0.0159,
      "step": 350
    },
    {
      "epoch": 1.46993006993007,
      "grad_norm": 0.6626363396644592,
      "learning_rate": 4.572079346027096e-06,
      "loss": 0.0099,
      "step": 351
    },
    {
      "epoch": 1.4741258741258743,
      "grad_norm": 1.0112382173538208,
      "learning_rate": 4.5687646003569055e-06,
      "loss": 0.0076,
      "step": 352
    },
    {
      "epoch": 1.4783216783216784,
      "grad_norm": 0.6920218467712402,
      "learning_rate": 4.565438277195265e-06,
      "loss": 0.0075,
      "step": 353
    },
    {
      "epoch": 1.4825174825174825,
      "grad_norm": 0.8465420007705688,
      "learning_rate": 4.562100395157379e-06,
      "loss": 0.0084,
      "step": 354
    },
    {
      "epoch": 1.4867132867132868,
      "grad_norm": 0.7478057742118835,
      "learning_rate": 4.5587509729231405e-06,
      "loss": 0.0129,
      "step": 355
    },
    {
      "epoch": 1.490909090909091,
      "grad_norm": 0.6042934656143188,
      "learning_rate": 4.555390029237026e-06,
      "loss": 0.0069,
      "step": 356
    },
    {
      "epoch": 1.495104895104895,
      "grad_norm": 0.7616173028945923,
      "learning_rate": 4.552017582907986e-06,
      "loss": 0.0103,
      "step": 357
    },
    {
      "epoch": 1.4993006993006994,
      "grad_norm": 0.7291771173477173,
      "learning_rate": 4.54863365280935e-06,
      "loss": 0.0139,
      "step": 358
    },
    {
      "epoch": 1.5034965034965035,
      "grad_norm": 1.2063847780227661,
      "learning_rate": 4.545238257878708e-06,
      "loss": 0.0064,
      "step": 359
    },
    {
      "epoch": 1.5076923076923077,
      "grad_norm": 0.37422144412994385,
      "learning_rate": 4.541831417117815e-06,
      "loss": 0.009,
      "step": 360
    },
    {
      "epoch": 1.511888111888112,
      "grad_norm": 0.6277711391448975,
      "learning_rate": 4.53841314959248e-06,
      "loss": 0.003,
      "step": 361
    },
    {
      "epoch": 1.5160839160839161,
      "grad_norm": 0.9660812020301819,
      "learning_rate": 4.534983474432458e-06,
      "loss": 0.0098,
      "step": 362
    },
    {
      "epoch": 1.5202797202797202,
      "grad_norm": 0.5692089200019836,
      "learning_rate": 4.5315424108313485e-06,
      "loss": 0.005,
      "step": 363
    },
    {
      "epoch": 1.5244755244755246,
      "grad_norm": 0.2833493649959564,
      "learning_rate": 4.528089978046481e-06,
      "loss": 0.004,
      "step": 364
    },
    {
      "epoch": 1.5286713286713287,
      "grad_norm": 0.731679379940033,
      "learning_rate": 4.524626195398812e-06,
      "loss": 0.0051,
      "step": 365
    },
    {
      "epoch": 1.5328671328671328,
      "grad_norm": 0.6426528692245483,
      "learning_rate": 4.521151082272817e-06,
      "loss": 0.0051,
      "step": 366
    },
    {
      "epoch": 1.5370629370629372,
      "grad_norm": 0.6782161593437195,
      "learning_rate": 4.5176646581163775e-06,
      "loss": 0.0075,
      "step": 367
    },
    {
      "epoch": 1.5412587412587413,
      "grad_norm": 0.8598656058311462,
      "learning_rate": 4.514166942440679e-06,
      "loss": 0.0119,
      "step": 368
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 1.1501063108444214,
      "learning_rate": 4.5106579548200935e-06,
      "loss": 0.0068,
      "step": 369
    },
    {
      "epoch": 1.5496503496503498,
      "grad_norm": 1.291435956954956,
      "learning_rate": 4.50713771489208e-06,
      "loss": 0.0066,
      "step": 370
    },
    {
      "epoch": 1.5538461538461539,
      "grad_norm": 0.6325899362564087,
      "learning_rate": 4.503606242357064e-06,
      "loss": 0.0046,
      "step": 371
    },
    {
      "epoch": 1.558041958041958,
      "grad_norm": 1.2787585258483887,
      "learning_rate": 4.5000635569783365e-06,
      "loss": 0.009,
      "step": 372
    },
    {
      "epoch": 1.5622377622377623,
      "grad_norm": 1.0738176107406616,
      "learning_rate": 4.496509678581936e-06,
      "loss": 0.0102,
      "step": 373
    },
    {
      "epoch": 1.5664335664335665,
      "grad_norm": 0.6169814467430115,
      "learning_rate": 4.492944627056544e-06,
      "loss": 0.0084,
      "step": 374
    },
    {
      "epoch": 1.5706293706293706,
      "grad_norm": 3.8595056533813477,
      "learning_rate": 4.4893684223533675e-06,
      "loss": 0.0061,
      "step": 375
    },
    {
      "epoch": 1.574825174825175,
      "grad_norm": 6.41073751449585,
      "learning_rate": 4.4857810844860325e-06,
      "loss": 0.0086,
      "step": 376
    },
    {
      "epoch": 1.579020979020979,
      "grad_norm": 12.252816200256348,
      "learning_rate": 4.482182633530468e-06,
      "loss": 0.0044,
      "step": 377
    },
    {
      "epoch": 1.5832167832167832,
      "grad_norm": 5.570111274719238,
      "learning_rate": 4.4785730896247985e-06,
      "loss": 0.0091,
      "step": 378
    },
    {
      "epoch": 1.5874125874125875,
      "grad_norm": 2.8494083881378174,
      "learning_rate": 4.474952472969224e-06,
      "loss": 0.0046,
      "step": 379
    },
    {
      "epoch": 1.5916083916083916,
      "grad_norm": 1.0975168943405151,
      "learning_rate": 4.471320803825915e-06,
      "loss": 0.0083,
      "step": 380
    },
    {
      "epoch": 1.5958041958041957,
      "grad_norm": 0.6368414163589478,
      "learning_rate": 4.467678102518894e-06,
      "loss": 0.01,
      "step": 381
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.6025326251983643,
      "learning_rate": 4.464024389433924e-06,
      "loss": 0.0088,
      "step": 382
    },
    {
      "epoch": 1.6041958041958042,
      "grad_norm": 0.8672336339950562,
      "learning_rate": 4.4603596850183915e-06,
      "loss": 0.0079,
      "step": 383
    },
    {
      "epoch": 1.6083916083916083,
      "grad_norm": 0.7727923393249512,
      "learning_rate": 4.4566840097811956e-06,
      "loss": 0.0077,
      "step": 384
    },
    {
      "epoch": 1.6125874125874127,
      "grad_norm": 0.7475773096084595,
      "learning_rate": 4.452997384292633e-06,
      "loss": 0.0089,
      "step": 385
    },
    {
      "epoch": 1.6167832167832168,
      "grad_norm": 0.5794192552566528,
      "learning_rate": 4.449299829184278e-06,
      "loss": 0.0062,
      "step": 386
    },
    {
      "epoch": 1.620979020979021,
      "grad_norm": 0.7462018132209778,
      "learning_rate": 4.445591365148875e-06,
      "loss": 0.0073,
      "step": 387
    },
    {
      "epoch": 1.6251748251748253,
      "grad_norm": 0.5339632630348206,
      "learning_rate": 4.4418720129402145e-06,
      "loss": 0.0036,
      "step": 388
    },
    {
      "epoch": 1.6293706293706294,
      "grad_norm": 0.4537692666053772,
      "learning_rate": 4.438141793373023e-06,
      "loss": 0.0049,
      "step": 389
    },
    {
      "epoch": 1.6335664335664335,
      "grad_norm": 0.9906306266784668,
      "learning_rate": 4.434400727322844e-06,
      "loss": 0.0072,
      "step": 390
    },
    {
      "epoch": 1.6377622377622378,
      "grad_norm": 0.48505276441574097,
      "learning_rate": 4.430648835725922e-06,
      "loss": 0.0045,
      "step": 391
    },
    {
      "epoch": 1.641958041958042,
      "grad_norm": 0.6711535453796387,
      "learning_rate": 4.426886139579083e-06,
      "loss": 0.0045,
      "step": 392
    },
    {
      "epoch": 1.646153846153846,
      "grad_norm": 0.6675395369529724,
      "learning_rate": 4.423112659939622e-06,
      "loss": 0.0097,
      "step": 393
    },
    {
      "epoch": 1.6503496503496504,
      "grad_norm": 0.5513535141944885,
      "learning_rate": 4.419328417925177e-06,
      "loss": 0.0027,
      "step": 394
    },
    {
      "epoch": 1.6545454545454545,
      "grad_norm": 0.5774904489517212,
      "learning_rate": 4.415533434713621e-06,
      "loss": 0.0045,
      "step": 395
    },
    {
      "epoch": 1.6587412587412587,
      "grad_norm": 0.39901164174079895,
      "learning_rate": 4.411727731542937e-06,
      "loss": 0.0021,
      "step": 396
    },
    {
      "epoch": 1.662937062937063,
      "grad_norm": 0.526216447353363,
      "learning_rate": 4.407911329711097e-06,
      "loss": 0.0077,
      "step": 397
    },
    {
      "epoch": 1.6671328671328671,
      "grad_norm": 1.100061297416687,
      "learning_rate": 4.404084250575952e-06,
      "loss": 0.0065,
      "step": 398
    },
    {
      "epoch": 1.6713286713286712,
      "grad_norm": 0.8158161640167236,
      "learning_rate": 4.4002465155551014e-06,
      "loss": 0.005,
      "step": 399
    },
    {
      "epoch": 1.6755244755244756,
      "grad_norm": 0.4041314423084259,
      "learning_rate": 4.39639814612578e-06,
      "loss": 0.0024,
      "step": 400
    },
    {
      "epoch": 1.6797202797202797,
      "grad_norm": 0.3537557125091553,
      "learning_rate": 4.392539163824738e-06,
      "loss": 0.003,
      "step": 401
    },
    {
      "epoch": 1.6839160839160838,
      "grad_norm": 0.6756988763809204,
      "learning_rate": 4.388669590248119e-06,
      "loss": 0.0044,
      "step": 402
    },
    {
      "epoch": 1.6881118881118882,
      "grad_norm": 1.2168368101119995,
      "learning_rate": 4.3847894470513356e-06,
      "loss": 0.0032,
      "step": 403
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 1.1175090074539185,
      "learning_rate": 4.3808987559489536e-06,
      "loss": 0.0036,
      "step": 404
    },
    {
      "epoch": 1.6965034965034964,
      "grad_norm": 0.3256933093070984,
      "learning_rate": 4.376997538714569e-06,
      "loss": 0.0029,
      "step": 405
    },
    {
      "epoch": 1.7006993006993008,
      "grad_norm": 0.9790744781494141,
      "learning_rate": 4.373085817180684e-06,
      "loss": 0.0067,
      "step": 406
    },
    {
      "epoch": 1.7048951048951049,
      "grad_norm": 1.3782650232315063,
      "learning_rate": 4.3691636132385885e-06,
      "loss": 0.0051,
      "step": 407
    },
    {
      "epoch": 1.709090909090909,
      "grad_norm": 0.7676941752433777,
      "learning_rate": 4.365230948838232e-06,
      "loss": 0.0075,
      "step": 408
    },
    {
      "epoch": 1.7132867132867133,
      "grad_norm": 0.4800433814525604,
      "learning_rate": 4.361287845988107e-06,
      "loss": 0.0031,
      "step": 409
    },
    {
      "epoch": 1.7174825174825175,
      "grad_norm": 0.5705649852752686,
      "learning_rate": 4.357334326755123e-06,
      "loss": 0.0023,
      "step": 410
    },
    {
      "epoch": 1.7216783216783216,
      "grad_norm": 0.6476736664772034,
      "learning_rate": 4.353370413264481e-06,
      "loss": 0.0035,
      "step": 411
    },
    {
      "epoch": 1.725874125874126,
      "grad_norm": 0.29256880283355713,
      "learning_rate": 4.349396127699552e-06,
      "loss": 0.0031,
      "step": 412
    },
    {
      "epoch": 1.73006993006993,
      "grad_norm": 0.44642290472984314,
      "learning_rate": 4.345411492301757e-06,
      "loss": 0.0025,
      "step": 413
    },
    {
      "epoch": 1.7342657342657342,
      "grad_norm": 0.43764063715934753,
      "learning_rate": 4.341416529370431e-06,
      "loss": 0.0024,
      "step": 414
    },
    {
      "epoch": 1.7384615384615385,
      "grad_norm": 0.47648465633392334,
      "learning_rate": 4.337411261262711e-06,
      "loss": 0.0044,
      "step": 415
    },
    {
      "epoch": 1.7426573426573426,
      "grad_norm": 0.7295253872871399,
      "learning_rate": 4.3333957103934025e-06,
      "loss": 0.0062,
      "step": 416
    },
    {
      "epoch": 1.7468531468531467,
      "grad_norm": 0.30801764130592346,
      "learning_rate": 4.329369899234857e-06,
      "loss": 0.0018,
      "step": 417
    },
    {
      "epoch": 1.751048951048951,
      "grad_norm": 0.40857842564582825,
      "learning_rate": 4.325333850316846e-06,
      "loss": 0.0054,
      "step": 418
    },
    {
      "epoch": 1.7552447552447552,
      "grad_norm": 1.0663392543792725,
      "learning_rate": 4.321287586226436e-06,
      "loss": 0.0042,
      "step": 419
    },
    {
      "epoch": 1.7594405594405593,
      "grad_norm": 0.7497619986534119,
      "learning_rate": 4.317231129607859e-06,
      "loss": 0.0024,
      "step": 420
    },
    {
      "epoch": 1.7636363636363637,
      "grad_norm": 0.6514341235160828,
      "learning_rate": 4.313164503162391e-06,
      "loss": 0.005,
      "step": 421
    },
    {
      "epoch": 1.7678321678321678,
      "grad_norm": 0.7208667993545532,
      "learning_rate": 4.309087729648217e-06,
      "loss": 0.003,
      "step": 422
    },
    {
      "epoch": 1.772027972027972,
      "grad_norm": 0.31625956296920776,
      "learning_rate": 4.3050008318803125e-06,
      "loss": 0.0027,
      "step": 423
    },
    {
      "epoch": 1.7762237762237763,
      "grad_norm": 0.38829004764556885,
      "learning_rate": 4.30090383273031e-06,
      "loss": 0.0039,
      "step": 424
    },
    {
      "epoch": 1.7804195804195804,
      "grad_norm": 0.8225162029266357,
      "learning_rate": 4.2967967551263725e-06,
      "loss": 0.0048,
      "step": 425
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 0.3218310475349426,
      "learning_rate": 4.292679622053066e-06,
      "loss": 0.0042,
      "step": 426
    },
    {
      "epoch": 1.7888111888111888,
      "grad_norm": 0.9199247360229492,
      "learning_rate": 4.288552456551229e-06,
      "loss": 0.0013,
      "step": 427
    },
    {
      "epoch": 1.793006993006993,
      "grad_norm": 0.6360734105110168,
      "learning_rate": 4.2844152817178476e-06,
      "loss": 0.0016,
      "step": 428
    },
    {
      "epoch": 1.797202797202797,
      "grad_norm": 0.1691218614578247,
      "learning_rate": 4.28026812070592e-06,
      "loss": 0.0027,
      "step": 429
    },
    {
      "epoch": 1.8013986013986014,
      "grad_norm": 0.503375768661499,
      "learning_rate": 4.276110996724332e-06,
      "loss": 0.0026,
      "step": 430
    },
    {
      "epoch": 1.8055944055944058,
      "grad_norm": 0.2944371998310089,
      "learning_rate": 4.271943933037726e-06,
      "loss": 0.0043,
      "step": 431
    },
    {
      "epoch": 1.8097902097902097,
      "grad_norm": 0.6249594688415527,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.0015,
      "step": 432
    },
    {
      "epoch": 1.813986013986014,
      "grad_norm": 0.7547388076782227,
      "learning_rate": 4.263580079886025e-06,
      "loss": 0.0032,
      "step": 433
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.32078686356544495,
      "learning_rate": 4.259383337227821e-06,
      "loss": 0.0021,
      "step": 434
    },
    {
      "epoch": 1.8223776223776222,
      "grad_norm": 0.4883444905281067,
      "learning_rate": 4.255176748478117e-06,
      "loss": 0.0016,
      "step": 435
    },
    {
      "epoch": 1.8265734265734266,
      "grad_norm": 0.22250854969024658,
      "learning_rate": 4.2509603371783776e-06,
      "loss": 0.0014,
      "step": 436
    },
    {
      "epoch": 1.830769230769231,
      "grad_norm": 0.5906175971031189,
      "learning_rate": 4.246734126925035e-06,
      "loss": 0.0042,
      "step": 437
    },
    {
      "epoch": 1.8349650349650348,
      "grad_norm": 0.2544255256652832,
      "learning_rate": 4.242498141369361e-06,
      "loss": 0.0012,
      "step": 438
    },
    {
      "epoch": 1.8391608391608392,
      "grad_norm": 0.6039454340934753,
      "learning_rate": 4.238252404217334e-06,
      "loss": 0.0042,
      "step": 439
    },
    {
      "epoch": 1.8433566433566435,
      "grad_norm": 0.7017889618873596,
      "learning_rate": 4.233996939229502e-06,
      "loss": 0.0039,
      "step": 440
    },
    {
      "epoch": 1.8475524475524474,
      "grad_norm": 0.3206017017364502,
      "learning_rate": 4.229731770220858e-06,
      "loss": 0.0008,
      "step": 441
    },
    {
      "epoch": 1.8517482517482518,
      "grad_norm": 0.2899838984012604,
      "learning_rate": 4.225456921060698e-06,
      "loss": 0.0008,
      "step": 442
    },
    {
      "epoch": 1.855944055944056,
      "grad_norm": 0.23223742842674255,
      "learning_rate": 4.221172415672492e-06,
      "loss": 0.0005,
      "step": 443
    },
    {
      "epoch": 1.86013986013986,
      "grad_norm": 0.41323575377464294,
      "learning_rate": 4.216878278033753e-06,
      "loss": 0.001,
      "step": 444
    },
    {
      "epoch": 1.8643356643356643,
      "grad_norm": 0.3661796450614929,
      "learning_rate": 4.2125745321758945e-06,
      "loss": 0.0003,
      "step": 445
    },
    {
      "epoch": 1.8685314685314687,
      "grad_norm": 0.577558159828186,
      "learning_rate": 4.208261202184104e-06,
      "loss": 0.0008,
      "step": 446
    },
    {
      "epoch": 1.8727272727272726,
      "grad_norm": 0.3006725013256073,
      "learning_rate": 4.203938312197203e-06,
      "loss": 0.0012,
      "step": 447
    },
    {
      "epoch": 1.876923076923077,
      "grad_norm": 0.40426135063171387,
      "learning_rate": 4.199605886407515e-06,
      "loss": 0.0026,
      "step": 448
    },
    {
      "epoch": 1.8811188811188813,
      "grad_norm": 0.7835168242454529,
      "learning_rate": 4.195263949060728e-06,
      "loss": 0.0038,
      "step": 449
    },
    {
      "epoch": 1.8853146853146852,
      "grad_norm": 0.32763028144836426,
      "learning_rate": 4.190912524455762e-06,
      "loss": 0.0011,
      "step": 450
    },
    {
      "epoch": 1.8895104895104895,
      "grad_norm": 0.12388205528259277,
      "learning_rate": 4.186551636944628e-06,
      "loss": 0.0002,
      "step": 451
    },
    {
      "epoch": 1.8937062937062938,
      "grad_norm": 0.29386594891548157,
      "learning_rate": 4.1821813109322975e-06,
      "loss": 0.0009,
      "step": 452
    },
    {
      "epoch": 1.8979020979020977,
      "grad_norm": 0.8328471183776855,
      "learning_rate": 4.177801570876561e-06,
      "loss": 0.0008,
      "step": 453
    },
    {
      "epoch": 1.902097902097902,
      "grad_norm": 0.35669347643852234,
      "learning_rate": 4.1734124412878915e-06,
      "loss": 0.0006,
      "step": 454
    },
    {
      "epoch": 1.9062937062937064,
      "grad_norm": 0.20081250369548798,
      "learning_rate": 4.169013946729312e-06,
      "loss": 0.0009,
      "step": 455
    },
    {
      "epoch": 1.9104895104895103,
      "grad_norm": 0.4637051820755005,
      "learning_rate": 4.164606111816256e-06,
      "loss": 0.0017,
      "step": 456
    },
    {
      "epoch": 1.9146853146853147,
      "grad_norm": 0.38190585374832153,
      "learning_rate": 4.160188961216423e-06,
      "loss": 0.0005,
      "step": 457
    },
    {
      "epoch": 1.918881118881119,
      "grad_norm": 0.13845530152320862,
      "learning_rate": 4.155762519649654e-06,
      "loss": 0.0001,
      "step": 458
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.09175657480955124,
      "learning_rate": 4.151326811887778e-06,
      "loss": 0.0002,
      "step": 459
    },
    {
      "epoch": 1.9272727272727272,
      "grad_norm": 0.37238427996635437,
      "learning_rate": 4.146881862754485e-06,
      "loss": 0.0011,
      "step": 460
    },
    {
      "epoch": 1.9314685314685316,
      "grad_norm": 0.08203031122684479,
      "learning_rate": 4.142427697125181e-06,
      "loss": 0.0006,
      "step": 461
    },
    {
      "epoch": 1.9356643356643357,
      "grad_norm": 0.4510544240474701,
      "learning_rate": 4.137964339926852e-06,
      "loss": 0.0004,
      "step": 462
    },
    {
      "epoch": 1.9398601398601398,
      "grad_norm": 0.23816390335559845,
      "learning_rate": 4.133491816137921e-06,
      "loss": 0.0019,
      "step": 463
    },
    {
      "epoch": 1.9440559440559442,
      "grad_norm": 0.42791134119033813,
      "learning_rate": 4.129010150788112e-06,
      "loss": 0.0041,
      "step": 464
    },
    {
      "epoch": 1.9482517482517483,
      "grad_norm": 0.677318274974823,
      "learning_rate": 4.1245193689583085e-06,
      "loss": 0.0026,
      "step": 465
    },
    {
      "epoch": 1.9524475524475524,
      "grad_norm": 0.21399132907390594,
      "learning_rate": 4.120019495780412e-06,
      "loss": 0.0002,
      "step": 466
    },
    {
      "epoch": 1.9566433566433568,
      "grad_norm": 0.16479253768920898,
      "learning_rate": 4.115510556437201e-06,
      "loss": 0.0008,
      "step": 467
    },
    {
      "epoch": 1.9608391608391609,
      "grad_norm": 0.28853413462638855,
      "learning_rate": 4.110992576162193e-06,
      "loss": 0.0002,
      "step": 468
    },
    {
      "epoch": 1.965034965034965,
      "grad_norm": 0.273203581571579,
      "learning_rate": 4.106465580239502e-06,
      "loss": 0.0001,
      "step": 469
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 0.2718943655490875,
      "learning_rate": 4.101929594003694e-06,
      "loss": 0.0019,
      "step": 470
    },
    {
      "epoch": 1.9734265734265735,
      "grad_norm": 0.33305594325065613,
      "learning_rate": 4.09738464283965e-06,
      "loss": 0.0012,
      "step": 471
    },
    {
      "epoch": 1.9776223776223776,
      "grad_norm": 0.05372636765241623,
      "learning_rate": 4.092830752182423e-06,
      "loss": 0.0002,
      "step": 472
    },
    {
      "epoch": 1.981818181818182,
      "grad_norm": 0.054683368653059006,
      "learning_rate": 4.088267947517088e-06,
      "loss": 0.0003,
      "step": 473
    },
    {
      "epoch": 1.986013986013986,
      "grad_norm": 0.5915653109550476,
      "learning_rate": 4.083696254378615e-06,
      "loss": 0.0017,
      "step": 474
    },
    {
      "epoch": 1.9902097902097902,
      "grad_norm": 0.9250974059104919,
      "learning_rate": 4.079115698351709e-06,
      "loss": 0.0008,
      "step": 475
    },
    {
      "epoch": 1.9944055944055945,
      "grad_norm": 0.22745700180530548,
      "learning_rate": 4.074526305070679e-06,
      "loss": 0.0003,
      "step": 476
    }
  ],
  "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": 1.2002884370772787e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}